# Qadvo > Qadvo keeps a dated record of each thing a business or a household owns and operates — equipment, vehicles, property, plantings — with what was done to it, what it cost, and what covers it, and turns that record into documents you hand someone. For landscapers, that document is the care guide: a guide to the customer's own yard that carries the landscaper's maintenance plan and sends the homeowner's requests back to the landscaper. Qadvo runs in a web browser on a phone or a computer and can be added to a phone's home screen. Facts reach a record by photo, voice, forwarded email, or spreadsheet import. Ask Q, Qadvo's AI assistant, answers questions from the account's own records and says so when a fact isn't in them. ## Who it is for - **Landscapers and landscape installers** (https://www.qadvo.com/; plan: Qadvo Business). After an install, Qadvo makes a care guide about the customer's own yard: each plant named, its care month by month for the zone, and the landscaper's own warranty terms and maintenance plan. The homeowner keeps it on their phone and on paper. A tap to report a problem or request work lands on the property's record and in the landscaper's inbox, and a Clients board follows each property's plan from offered to active. Why it matters: after an install, the maintenance plan and the next job usually leave with the truck. The guide keeps the landscaper's offer in the customer's hands and brings the customer's requests back — without writing each guide by hand. - **Farms, shops, and small fleets** (https://www.qadvo.com/features; plan: Qadvo Operation, or Qadvo Business for a team with customers). Each machine and vehicle keeps its own dated record of service, fuel, repairs, and readings, with the receipt or invoice attached. From it come what each one costs per hour or per mile, service reminders by hours, miles, or dates, the vehicles and equipment no policy on file covers, and a dated history for a warranty claim, an insurance claim, or a sale. Decisions it informs: whether to repair, replace, or keep a machine, on its own numbers; what to ask the insurance agent about; what is due next. - **Households** (https://www.qadvo.com/?variant=home; plan: Qadvo Personal). One trustworthy record of the home — the largest thing most people own — and everything else they own: the cars, and what is in every room. Receipts, warranties, policies, and every repair, dated, with the paperwork attached. Five photos of a room — one of each wall and one of the whole room — become a draft inventory with estimated replacement values (beta), checked by the owner and printed as an Insurance Inventory report. Ask Q answers from those records, not from the internet, and says when the answer isn't in them. Family members view for free, and a record can go with an item when it is sold. The problem it is built for is the open loop a homeowner carries — not knowing what they have, what it needs, or whether they are behind — so the record takes the house as it is, unpacked boxes and all, and is built for a home that keeps changing. Q gives the facts the record holds; the decisions stay the owner's. Decisions it informs: whether a repair is already covered before you pay for it yourself; what is actually due and what can wait; what to show an insurer, a buyer, or a contractor. ## What Qadvo is not - Not fleet telematics: no GPS tracking, ELD or hours-of-service logs, dash cameras, or fuel cards. Hours, miles, and fuel are logged by people — by photo, voice, a forwarded receipt, or an import — not read from the vehicle. - Not field-service management: no job scheduling, crew dispatch, routing, estimates, or invoicing. For a landscaper's maintenance plan, Qadvo records the agreement and its status; the customer is invoiced however the landscaper invoices today, and Qadvo does not take that payment. - Not bookkeeping: it produces an expense and tax-preparation export (PDF and spreadsheet) for an accountant, not a set of books. - Connections: forwarded email in, spreadsheet import, and PDF and spreadsheet export. There are no direct integrations with other apps yet (a SynkedUP connection through Zapier is planned). ## Where it fits Qadvo sits beside the software a business already uses to schedule, estimate, and invoice — for a landscaping company, job software such as SynkedUP, Jobber, LMN, or Aspire — and beside telematics on vehicles that have it. It keeps the record of each thing and hands it to someone; it does not replace those tools. It has no direct connection to any of them yet; a SynkedUP connection through Zapier is planned. ## Pricing (USD) - **Qadvo Personal** — $119/year. For one household's things. - **Qadvo Operation** — $499/year, or $49 a month, month to month. For a farm, a shop, or a one-truck business. - **Qadvo Business** — $990/year, or $99 a month, month to month. For a business with a team and customers. Every plan starts with a 45-day free trial: a card is required, nothing is charged until day 46, and it can be canceled anytime. Viewers are free and unlimited on every plan. Full comparison: https://www.qadvo.com/pricing ## Status Qadvo is available now on all three plans; care guides are part of Qadvo Business. It is updated often. These facts describe Qadvo 1.52.66, released 2026-10-18. ## Pages - [For landscapers: the care guide](https://www.qadvo.com/): the front page — what a care guide is, how it is made, and the Business plan - [For households](https://www.qadvo.com/?variant=home): Qadvo for a home, its vehicles, and its belongings - [About Qadvo](https://www.qadvo.com/about): these facts as a web page: who it is for, what it is not, pricing, and status - [Everything Qadvo does](https://www.qadvo.com/features): the full feature list, one plain line each - [Pricing](https://www.qadvo.com/pricing): the three plans, what every plan includes, and how they differ - [An example care guide](https://www.qadvo.com/report/care-guide/view/1f2a7cea-ad94-4b2c-971f-57fca491d305): a real published guide, as the homeowner sees it - [Maintenance guide](https://www.qadvo.com/guide): a free guide to which maintenance is worth tracking, and why ## Optional - [Full reference](https://www.qadvo.com/llms-full.txt): this file, the full feature list, the plan comparison, and the user manual that Ask Q, Qadvo's assistant, answers from - Contact: admin@qadvo.com ## Everything Qadvo does ### Capture Get a fact onto the record in seconds, from wherever it is. - Photograph a receipt, a nameplate, a label, a manual page, or a whole room — Qadvo reads it: the fields fill and the file stays on the record as proof. - Forward an invoice or order confirmation by email; it lands in your inbox as a card and files to the item it names — or asks you which, when two could fit. - Say it: voice logging turns "changed the oil on the Kubota at 412 hours" into a dated entry. - Scan a QR sticker on the machine and log against it without searching. - A shop whiteboard, a card statement, or a dated sessions sheet becomes many entries at once — the table reader. - Bulk history: drop a folder of old invoices and each becomes a review-ready entry with the original attached. ### Records and timelines Every item keeps its own history. Nothing is erased. - One record per item — equipment, vehicles, property, plants, hardscape, documents — nested any depth (a farm → a truck → its trailer). - A timeline of every service, repair, fuel-up, reading, incident, and improvement, with the evidence photo or PDF attached. - Coverage and documents: policies, warranties, manuals, and contracts live on the item they cover; a fleet policy on the group covers every truck under it. - Item status: sold, traded in, scrapped, lost, retired — the ledger closes honestly instead of disappearing. - A contents ledger for a room, a shed, or a trailer, built from a photo, for insurance or a move-out (beta). ### Reminders Service intervals that reset when you do the work. - By hours, miles, acres, cycles, or dates — oil every 250 hours, inspection every April, warranty ending in 14 months. - Compound intervals (5,000 miles or 6 months, whichever first) and custom warning thresholds. - A Due list on the home screen and an email reminder before the date, both scoped to your own work. ### Ask Q Ask your record anything. Answers come from your data, not the internet. - "What did the F-150 cost me last year?" "Which warranty expires first?" "When did we last service the sprayer?" - Q reads the item you have open, its documents, and its history — and says plainly when a fact isn't in the record. - Ask Q to set up a reminder and it proposes the cycle and start date from your record; you confirm before anything is saved. ### Care guides A document about your customer's own yard, with your offer on it, that writes back to you. - Photograph the install; the guide names each plant for you to confirm and writes its care, month by month, for your zone — watering amounts, protect-it notes, what looks alarming but isn't. - Publish it: a frozen, dated edition with a link and a QR code, on the customer's phone and on paper, carrying your maintenance plan and your own warranty terms. - Something wrong with a plant? They tap it. The request lands on the property's timeline and in your inbox, with their note. - A per-property service agreement with tiers, offered on the guide and tracked from offered to active; a Clients board of every property, its plan, and what's due. ### Reports The strongest thing a record can do is prove something to someone else. - Cost to own: what each machine actually costs per hour or per mile, from what you logged. - Expense and tax export: every business expense by item and type, as a PDF and a spreadsheet. - Certified History: a frozen, dated, shareable record of one item — every entry, every warranty, for a sale, a claim, or an audit. - Insurance inventory and rental condition reports, built from a room's contents ledger (beta). - Lab calibration reports for instruments, and a crop-year analysis for grain operations. ### Sharing and client work Your records, in the hands that need them, at the level you choose. - Share any item or group as viewer, editor, or admin; children inherit the parent's share. Viewers are free. - Transfer ownership when you sell — the history goes with the item. - Client work stays separate from your own: one pill flips the home screen between Own and Clients; each client property can carry a maintenance account and its own service agreement. - A customer who receives a care guide can claim a living record of their property and, with edit access, add photos to the same record you work from. ### Value and economics What it's worth, what it cost, and what it would take to replace. - Purchase price, current value, replacement cost, amount owed, and net value on every item; a portfolio total across everything you own. - AI value estimates, and a Sell This Item flow that drafts the listing and researches an asking price. - Depreciation and cost of ownership for anything marked a business expense. ### Security and privacy Your data is yours. - Private storage with row-level access rules; shared links are unguessable tokens; published reports are hidden from search engines. - AI runs through private API connections — nothing you log trains a model or is visible to anyone else. - Export a full backup any time. Delete your account yourself from Settings; after a 30-day recovery window, everything is purged. ## The plans in detail ### Qadvo Personal — $119/year For one household's things. - Every item, receipt, warranty, and service in one place - Timers and the Due list — a reminder before a service interval or renewal comes due - Insurance inventory, condition, and history reports - Ask Q anything about your records - Share with family — viewers are always free ### Qadvo Operation — $499/year, or $49 a month, month to month For a farm, a shop, or a one-truck business. - Everything in Personal - Business flag and the tax-preparation report - Your business profile on every report - Imports, batch actions, and status for the whole fleet or herd - One seat — a second set of hands with their own login - Included AI — three times Personal's ### Qadvo Business — $990/year, or $99 a month, month to month For a business with a team and customers. - Everything in Operation - Care guides with your plan and warranty on every one you publish - Customer requests land on the property's timeline and in your inbox - Five seats for your crew and office; viewers always free - Included AI — ten times Personal's - Two 1-on-1 onboarding sessions — we'll get your records loaded together. What every plan includes: - Every item, receipt, warranty, and service on one timeline — nothing is erased unless you delete it - Timers by hours, miles, or dates, the Due list, and reminder emails - Insurance inventory, condition, and certified history reports - Ask Q — questions answered from your own records - Photo, voice, forwarded-email, and spreadsheet capture - Item status (sold, retired, replaced…) and batch actions - Sharing with free viewers — always - Records transferred to you are yours forever, plan or no plan | | Personal | Operation | Business | | --- | --- | --- | --- | | Built for | one household's things | a farm, a shop, a one-truck business | a business with a team and customers | | Seats for people who work in the account | owner only | 1 included | 5 included | | Viewers | free, unlimited | free, unlimited | free, unlimited | | Included AI each month (fair use — a meter in Settings shows where you stand) | included | three times Personal | ten times Personal | | Business flag and tax-preparation report | — | ✓ | ✓ | | Your business profile on reports | — | ✓ | ✓ | | Care guide selling mode, customer requests, branded link | — | — | ✓ | | 1-on-1 onboarding sessions | — | — | two | A seat is a person who works inside your account with their own login — invite them in Settings → Team. Viewers are free on every plan. The line between Operation and Business: reports you use to run your business are Operation; documents you hand a customer are Business. Every plan starts with a 45-day free trial: a card is required, nothing is charged until day 46, and it can be canceled anytime. ## The user manual Below is the manual Ask Q, Qadvo's assistant, answers how-to questions from — the same text, so it describes the app as it ships. It is written for people using Qadvo; version tags such as (v1.51.152) mark when a feature arrived. ### Qadvo — how the app works (reference for how-to questions) #### Capture: logging, scanning, voice — the "+" menu in Ask Q The fastest way to capture anything is the **"+" button inside the Ask Q box**. Tap it for: - **Scan** — opens the camera to scan a QR sticker on an asset; you land in Ask Q scoped to that asset, ready to log what you did. - **Speak** — hands-free voice: say what you did ("changed oil on the truck, $45") and Q turns it into a structured log. **Which item it lands on:** from the general Ask Q, name the item — if Q can't tell which one you mean, it asks you to pick. From Ask Q on an OPEN item, there is no need to name it: the log goes to that item (the Speak panel shows "Logging to " before you talk), or to something inside it if you name that instead. Q only asks there when you name something inside the open item that matches more than one thing. **Shared items and teams:** Speak follows the same permissions as +Log — it works on your own items, on items shared with you as editor or admin (and everything inside them), and, for a team member, on the workspace owner's items. Items shared with you view-only are never offered as a voice-log target. Every voice log records who entered it, so inside a workspace the owner sees which team member logged it. **Stopping (v1.52.39):** tap Stop and listening ends at once — the button turns to "Processing…" and what you said goes through, even when the phone hadn't confirmed the words yet; or just stop talking: listening ends on its own after **5 seconds of silence** (pauses shorter than that keep listening). Before the first words there's longer — **8 seconds from when the phone starts listening** (v1.52.40) — and with nothing said it stops and says "Didn't catch anything". Before v1.52.39 an iPhone could turn "Listening… Tap to Stop" back on after Stop and keep the microphone open. **Review before saving (v1.52.40):** a switch in the Speak panel, on unless you turn it off and remembered on that device. With it on, nothing is saved until you've checked it: the item, the log type, what happened, the cost, the date and each amount are all editable, then **Save log**. With it off, the log saves as soon as listening ends. **Only numbers you said are used (v1.52.40):** an amount or a cost you didn't say is left blank, never guessed — the review shows it in amber ("You didn't say Gallons…"), and with review off the log saves without it and the Inbox note says it was left out. Before v1.52.40 a fill-up with no amount said could be logged as "1 gallon". **Which item (v1.52.41, v1.52.42):** when what you say fits more than one item ("the truck" with two trucks), Q doesn't guess: a reading you say decides between them — a meter only counts up, and only so fast (at most 1,000 miles or 24 hours a day since its last reading). A round number counts as rounded: "48,000" fits a truck that reads 48,310. A reading that fits none of them picks the one whose last reading is clearly nearest (at least 4× nearer than the next). The review says why ("Picked by the reading: … fits , not ", or "Picked as the closest: …"). If the reading fits several, none is clearly nearest, or nothing you said tells them apart, the item is left blank with those items first in the list, and you pick. Only meters decide — a timer's own length or a gallons amount never does. A meter reading said with any kind of log — a fill-up "at 48,000 miles" as much as an oil change — is kept on the item's meter and counts for picking the item (v1.52.44; fill-ups had been losing it). A reading below the item's last one is flagged on its field ("below the last reading — if you rounded, type the exact reading"); a log that saves without review is flagged in the inbox. The Speak panel is its own sheet over the page (it scrolls inside itself) and closes only with its X, so a tap outside — like the one that puts the keyboard away — no longer throws the log away. **Meters in the review (v1.52.42):** the review always offers the item's meters — the odometer or hour meter its timers count by — blank, with the last reading as a hint, even when you didn't say one. A log that restarts a timer ("oil change" restarts Oil Change) says so, and its meter is marked: type today's reading and the timer restarts from it. Left blank, the timer restarts from the last reading only if that reading is from the past week; an older one would restart it from a stale point, so the timer is left alone and you're told. A reading you didn't say is never written onto the log. **When to talk (v1.52.42):** the button reads **Opening the mic…** until the phone is actually hearing, then **Speak now** — on an iPhone the microphone opens a second or two after the tap, and words said before that are lost (the whole try, not just the first words). A try that hears nothing says which: "The microphone didn't open" (wait a moment, then tap again) or "Didn't hear anything" (wait for Speak now). On iPhone and iPad the Speak panel no longer talks back: the page speaking switched the phone's audio away from the microphone, so the next tries heard nothing. Its messages are on screen only. **Speak records (v1.52.43):** Speak now records your voice and Qadvo's server turns the recording into words — the same review and save follow. An iPhone's built-in speech recognition isn't supported by Apple in apps added to the Home Screen: there it heard the first try after the app opened and nothing after until the app was closed. Recording works every try, in the Home Screen app, Safari, Android and computers. A green bar shows the sound it's hearing; it stops 5 seconds after you stop talking (or tap Stop), and never runs past a minute. Turning the recording into words takes a second or two after it stops. The phone asks for the microphone the first time (a Home Screen app may ask again after it's reopened). If the bar can't show — the phone didn't let the page read the sound — it says "Tap Stop when you're done" and stops only on Stop. A browser that can't record still uses its built-in recognizer. **After saving (v1.52.45):** a note says where the log went ("Logged to ") with a **View log** button that opens that item on its Activity tab — it never jumps there on its own, since voice logs often come in runs. In the review, the "You didn't say …" note lists only amounts still blank: typing one in clears it. - **Photo** — opens the camera. Each shot stages in a tray so you can snap SEVERAL angles (tap "+ Add another"), then "Analyze" submits them together — different views of the same thing merge into one item (a close-up plus a wide shot gives the best plant/product ID). - **Upload a photo** — pick images already on your device (screenshot, saved receipt, gallery photo); you can select several at once and Qadvo merges views of the same subject into one item. Qadvo identifies what it sees: products and serial plates, vehicles by VIN, animals (breed, sex, estimated weight), and **plants** — snap or upload a photo of a tree, shrub, vine, or houseplant and Qadvo identifies the species (common and botanical name), sun needs, mature size, and a key care fact. A leaf close-up plus a whole-plant shot together give the best identification. You can also just type the action to Ask Q ("changed oil, $45", "log 10 gallons at 46,500 miles") and Q creates the log for you to confirm — the confirmation shows every number Q read (mileage or hours, gallons, cost) as editable fields next to the Yes button, so a misread value can be corrected in place before saving. Include the meter reading in your message and it's one tap. To log from an asset directly: open it → **Activity** tab → **+ Log** → pick the event tab (Maint, etc.) → add details → Save. #### Importing logs from a spreadsheet Open an item → **Activity** → the import option (or the bulk log importer) lets you map spreadsheet columns to log fields and import many entries at once. Two things it does for you: (1) a row that carries only readings/measurements (numbers like weight, a score, a rating — no cost, no vendor, no work description) is filed as a **Measure** (measurement), not Maintenance — so imported metrics match how the app logs them by hand. (2) Before importing, it checks each row against what's already on the item and flags **duplicates** (same date and same readings); duplicates are **skipped by default**. The preview shows how many are new vs. already-there and lets you decide per row — tap a duplicate's chip to skip or import it, or use the bulk "Skip all" / "Import all anyway" buttons. So re-importing an overlapping sheet won't create double entries unless you choose to. Long imports page in with a "Show more" button rather than stopping at the first handful. #### Import Receipts & Records (invoices) — when a scan fails or stalls (v1.51.152) Edit Details → **Import invoices** (or the +Log paths that use it) scans each file separately and shows a review screen before anything is written — nothing lands on the item until you click **Import** there. If a file's scan request fails (a dropped connection, a gateway timeout, a server restart mid-run) it is **never dropped quietly**: the review screen opens with an amber banner naming the files that didn't make it, with **Retry these** and **Skip them**. Each scan is cut off after two minutes and retried once automatically (a "Retrying…" note shows under the progress bar), so one hung file can't stall the batch. While it runs you can **Stop and continue with the N read so far** (takes what came back to review and lists the rest for a retry) or **Cancel import** (stops everything, keeps nothing, back to the upload step). Closing the window is the same as Cancel. Rows that are pages of one paper are merged and show a "N pages merged — Split" badge; a file that couldn't be read still appears as a row flagged for manual entry. #### Deleting a log entry Open the asset's **Activity** tab, find the entry, tap the small trash icon next to the pencil — it expands into a red "Tap again" pill; tap once more within 3 seconds to confirm. (Or tap the pencil to open the edit modal, which has a labeled Delete button.) #### Deleting an item Open the item → **Edit** (pencil) → scroll to the bottom → **Delete This Item** — it turns red with "Tap again"; tap once more within 3 seconds to permanently delete. Owner-only. **Deleting a group or parent never deletes what's inside it**: its children move UP one level (to the deleted item's parent, or the home screen) — so deleting a wrapper group simply dissolves it. To delete contents too, delete them individually first. Deleting an item deletes its files, except a file another record still shows (two items made from one paper each keep it). EXCEPTION: the read-only archived copy of an item you transferred away has no Edit button — for those, a **Trash** button appears directly in the item's action row (same two-tap confirm). #### Adding a new item Tap **+ Add Item** on the home screen. The window ("Add to Main Vault", or "Add to " when you add inside one) offers **Create from a Photo or Document** (upload a receipt, label, manual or any image; the AI extracts the details, and **Tell the AI what it's looking at** is optional), **Create Item Manually**, and **Create from Spreadsheet** for bulk. At the top level it also shows **Or create from email**: your save address with a copy button — forward a receipt, invoice or notice there and it's logged to the item it belongs to, or the item is created if it's new (see Forwarded receipts). Inside an item it offers **Or move existing item** instead. Fastest of all: tell Ask Q ("add my Toyota Tacoma") and Q sets it up. **Reviewing a scanned receipt (v1.52.34):** each line Q read is a card with its name, price, store and date; everything else (type, brand, model, year made, serial, payment, tax, notes, quantity, business) is under **More details**. A line's name is what you'd call the thing; the receipt's own wording (codes, options, discounts, order number) goes in its notes, so nothing is lost. The top of the review shows the receipt's printed subtotal, tax and total and whether the lines add up to its subtotal. **Combine items:** when a receipt lists parts of one thing on separate lines (a frame and its lenses, a mower and its deck), tap **Combine items**, tick the lines and tap **Combine N into one item** — any two or more, as many groups as needed. The price is the lines' sum, each line is listed in the notes, and **Split** undoes it; give the combined item its own name before saving. A combined item keeps each part's own maker's warranty as its own timer, named for the part ("Lenses warranty"), listed by part on the Coverage tab; a part whose brand isn't on the receipt gets none (fix the brand under More details first). **Protection plans:** a protection plan or extended warranty on the receipt shows inside the card of the item it protects and is saved as that item's coverage — a Warranty / Protection Plan document filed under it, starting on the purchase date and ending after its length when the receipt prints one (otherwise type the length under More details). It stays unless you tap **Remove**. If Q took a plan for an item, open that card's More details and pick **A protection plan? Yes, it covers …**; the other way, a plan's **Not a plan? Make it its own item**. An add-on bought for another line (a bagger for its mower) is filed under it and says so, with **Keep separate**. **Q read something wrong? Re-scan with a note** re-reads the paper with your note; the note you typed when starting the scan is already in the box. Prices, tax shares and dates are worked out by the app from the printed numbers, never by the AI: a line's own discount is taken off once, and a date is saved as a date whatever its printed form. Documents filed under an item (its plan, manual or title) are not counted in its Contents value. #### Fuel-up by photo (pump capture) Log a fuel-up without typing: from Ask Q's **+ → Photo**, snap the pump's display (or the printed fuel receipt) and your dashboard odometer — shots stage in a tray so you can snap the dash while pumping and the display after; tap Analyze once. You can type a note with the photos ("fuel up the old chevy") to name the vehicle. Qadvo reads the $ total, gallons, price per gallon, and odometer — and from a receipt also the transaction date (the receipt's own date, not scan day), station name, and tax, so business records stay complete. A review opens before anything saves: the target vehicle is shown and changeable, and **every field is editable** (odometer, gallons, cost, date, vendor, tax, notes — default note "Refuel"). Vehicle identity: a name or tag number on the gas door — a Qadvo tag, a label-maker sticker, or handwritten marker — is read from the photo and matched against your items including those shared to you with editor access; a fuel log can only target vehicles/equipment, never a policy or group. No name visible → pick from a list. **Photos are discarded by default** (pump screens and dashboards are transport, not records — a receipt in the batch defaults to KEPT as evidence); each photo has its own Keep checkbox, and canceling the review deletes the uploaded photos immediately. After saving, the app opens the vehicle's Activity tab showing the new entry. #### Importing items from a spreadsheet (bulk add) **Batch three (v1.51.134):** +Log (photo or PDF) handles a loan/lease statement the same way Import does — Record, no cost, Loan Balance metric, Amount Owed updated. **How to log a loan or lease payment (the one rule, v1.51.155):** scan the statement — Import Receipts & Records, or +Log → Scan — and it lands as a Record with the principal, interest, balance, and payment stored as figures and NO cost. That is deliberate: the principal is the purchase price paid in installments and is already in the item's value, so it is never an expense anywhere; the interest is the expense, and the tax report picks it up from those figures. Don't type a loan payment in as a cost log — that counts the whole payment as spend, overstating Economics and the deduction. The typed log form has no statement fields yet; scan the paper instead. +Log warns before saving a log that looks like one already on the item (same statement/invoice number, or same date plus vendor, total, or odometer) and hides Save until you confirm it's a separate visit. Monthly statements never merge into one record (an account number is not a statement number). A GROUP's own logs — fleet-wide receipts like one service visit covering seven snowblowers — count in the group's Economics as a "Shared — logged on the group" row (not split per unit), and a group's Edit screen offers Import Receipts & Records and Import Maintenance Logs. The group's Economics table lists EVERY member at any depth — a calf nested under its mother, an attachment under a tractor — indented under its parent; each row carries only its own value and logs (nothing is counted twice), and the totals and the "N of M tracked" count cover the whole tree. "Surprise me" spotlights an item with real history and tells you something its own logs don't already say (see **Surprise me** under Ask Q itself). Hovering a log's thumbnail shows the file name. **Lender statements and multi-page invoices (v1.51.131):** the invoice importer (Add Item → Import, or Log → scan on an item) recognizes a monthly LOAN or LEASE statement from a finance company: it becomes a Record with the PDF attached, the payment, principal, interest, balance and due date in the description, and NO cost — principal is the purchase price arriving in installments (already counted once) and interest is a cost of owning rather than running the item, so cost per hour is untouched; a "Loan payment · no cost" chip marks these rows. The NEWEST statement sets the item's Amount Owed (the one scan allowed to, because it is the lender stating today's balance — previewed as an "Amount Owed →" chip before import), and the balance is a tracked metric (Loan Balance) so the payoff curve shows in the Performance Monitor. Ask Q "what do I owe on the 289?" reads Amount Owed; "how much interest this year?" reads the Records. An invoice scanned as several files is merged into ONE log when the pages share an invoice number, page marks, a scanner file name like "Inspection 3of8" / "8of8", or vendor + date (v1.51.132: the grouping runs over the whole batch, and every merged page is listed on the review row, each openable) — the page with the total leads, every PDF is attached, and a "N pages merged — Split" chip undoes it. Duplicates: the paper's own invoice/statement number is the first signal, so re-importing a folder skips what is already logged. A parent's economics never counts QDocs (policies, receipts) under it. **+ Add Item → Import** accepts .xlsx, .xls, and .csv — and from inside any group, **Add Content → from this item's photos** (or any photo upload) opens the **Review Found Items** screen once the AI has read the photos — that wait is the reading, not the screen. Each item shows **Photos of this item**: the first is its picture (marked Cover); tap a photo to make it the cover, the small box takes a photo out of (or back into) that item, the magnifier enlarges it — a plant is identified from a big picture. The scan guesses which photos show which item; correct it there before Create, or every item keeps every photo (v1.52.14). **Add Content → Import** imports INTO that group (everything created lands inside it, not at the root — how a whole business schedule stages under one bucket). Two ways: download the **Qadvo template** (one tab per item type — every column maps automatically), or **use your own spreadsheet** — often easier for large lists. Own-spreadsheet setup: column headers in the **first row**, one row per item. A **Name column is required**; common headers (Name, Make, Model, Year, VIN, Serial, Price, Value, Location, Notes, Sub Category, Quantity, Ledger Parent/Container) map automatically; after upload you match or skip the rest by hand — nothing imports without review. One tab imports at a time; for big lists use **one flat tab with a Type column** (the 7 types — Vehicle, Equipment, Item, Property, Life, QDoc, Group — or your own words, matched once per word). **Records vs Ledger (desktop import step):** each row imports as either a full **Record** or a counted **ledger line** on a container. A REAL serial/VIN or license plate forces Record (warranty tracking needs a serial); everything else splits on a price threshold you set (default $250, under → ledger). Every row shows in a full-width table (resizable columns, pinned header) with a one-tap Record/Ledger flip. **When in doubt choose Ledger — ledger lines promote to full records in one tap later; the reverse isn't possible.** A **Ledger Parent / Container column** names which group/property each ledger row's lines land on (values map once, containers are created as needed inside the import parent); a **Sub Category column** becomes the ledger line's *area* AND the record's sortable Sub-Category. **Records group into Sub-Category buckets automatically** (a bucket is only created for 3+ records; smaller sets sit directly on the parent) — the group tree is mobile's navigation, so a big import arrives organized, not as hundreds of flat siblings. Imported values and counts are stamped as OWNER facts — camera scans verify them but never overwrite them; imported lines start unverified, and the first room scan is the verification pass. The import screen shows live progress; large schedules take several minutes. #### Service intervals & reminders (timers / countdowns) **Intervals** are recurring service reminders (e.g. oil change every 5,000 miles or 6 months). **Countdowns** are one-time windows like a warranty that don't reset. Set them from an asset's **Activity** tab (the **+** / Timers in the Performance Monitor), or just ask Q to "set up an oil change reminder every 5,000 miles" and it creates it. They track automatically and alert when due. Intervals apply to an individual asset, not a whole group — set them per item. To change a cycle, tap the interval card on the asset's Activity tab to reopen the wizard pre-filled. A timer's name is one name however it is capitalized — re-saving a timer with different capitals keeps the one timer (v1.52.48). **Starting Point — Start fresh / Already serviced (v1.52.30):** editing an interval (tap it → Edit Timer) shows where it counts from now, and the choice you save is the word on it — it outranks every service log already on the item. **Start fresh** clears the count: a miles/hours interval shows "Needs first log" until the next service is logged (a calendar interval counts from the day you saved it). **Already serviced** counts from the date and reading you give; tapping it fills in the last service on record — the date and meter reading the timer counts from now — ready to check and correct in one place (v1.52.31). A service logged AFTER you save — even one dated earlier, like last month's receipt entered today — is newer information and restarts the count as usual. The Activity tab, the timer's detail view, the Due list, Overview → Maintenance, the reminder emails and Q all read the same starting point. Before this, "Start fresh" on an interval that already had a logged service (an imported invoice, say) changed nothing: every screen kept counting from that old log. Since v1.52.33, a timer set up before starting points existed (no Start fresh / Already serviced saved on it) restarts the same way: a service logged after its set-up date counts from the service's own day, even when the set-up carried a start date of its own. Since v1.52.33, the dashboard card's first line reads the same rules: it shows a timer exactly when the Due list counts it (warranties included), else the item's current reading (the odometer or hours from the newest log); a group's card shows the newest log anywhere inside it, named for the item it's on. A timer's name is saved trimmed, so a stray space can't split one timer into two. **Compound warranties — "24,000 miles OR 24 months, whichever comes first" (v1.51.121):** the timer wizard's "Add second condition" makes ONE timer with two bars in the Performance Monitor (e.g. a miles bar and a months bar). Every warranty now has a Start Date (the install/purchase/replacement date), including mileage-based ones — set it so the time half counts from the right day. Whichever limit is reached first expires the whole timer; archiving or deleting the timer removes both bars. Q can create a compound warranty from chat too ("set up a 24,000-mile or 24-month warranty on the converter from Nov 24, 2023"). **Two timer toggles, two jobs (v1.51.121):** "Show on Dashboard" controls whether a timer appears in the home screen's Due list and counts toward an item's red/amber badge; "Send email reminder" controls only whether it is included in reminder emails. Turning emails off no longer hides a timer from the Due list (it used to). **Alert threshold (when the warning fires):** each timer — intervals AND warranty countdowns — has an "Alert at X% used" slider (50–95%, default 90% since v1.51.123 — about a month early on a year-long timer; earlier timers saved with 75% keep it until edited) under its notification settings — in the timer wizard and in the metric's detail view. Under the slider a live line translates the % into concrete terms ("≈ 250 hours remaining when alerted", or "≈ 3 months before due" for a warranty), and each timer's progress bar shows a small amber tick marking exactly where the warning zone begins. The threshold controls when the bar and Due list turn amber for that timer. **Email reminders default ON** for every timer (including auto-created warranties) unless the user unchecks "Send email reminder" on that timer. #### "Due" filter & what's due The **Due** filter pill on the home screen shows assets with upcoming or overdue maintenance intervals and expiring warranties (red badge = overdue count, amber = due-soon). For a sortable cross-asset view, open **Overview → Maintenance**. You can also just ask Q "what's due?" or "what's my top priority?" for a live roll-up with one-tap "Log it" buttons. **Muting reminder emails for a handed-off property:** if you've shared a property with a customer who now cares for it, open its **Share** screen and tap **"Handed this off? Mute its reminders for you."** You'll stop getting maintenance reminder *emails* for that property and everything nested under it — but it stays in your own **Due** list, and the customer keeps their view. It's per-property, off by default, and one tap turns it back on if you still service the property. #### Home-screen filter pills **Return window and warranty lookups show their proof (v1.52.50):** when Qadvo looks up a purchase's return window or its maker's warranty, it searches the seller's own website first — the site the receipt email links to — and then the wider web, and it keeps a term only when the page belongs to the right business (the seller for a return window, the maker for a warranty) and states the term in its own words. The timer's note carries that sentence and its link goes to that page; a search-engine link is never a source and there is no "usual" figure — when nothing can be backed, no timer is set and the inbox note says it couldn't be verified. Businesses that share a name (a store, a school and a factory under one brand word) no longer cross over. A lifetime warranty, or one measured in miles or hours, has no end date: it is shown as terms on the Coverage tab and nothing counts down. Each lookup takes about ten seconds. **Re-scan email (v1.52.50):** an item created from a forwarded email keeps the email as its receipt; open it from the item's "Receipt on file" tile (or its photos) and tap the orange **Re-scan email** button. Qadvo reads the email again (filling the maker or model when the record has none), looks the return window and the warranty up again, and brings the timers IT had looked up in line: corrected, confirmed, or removed when they can't be backed — the message that follows says which and why, and stays until closed. Timers you set yourself, a return window read off a paper receipt, a seller's promise and a card's extended-warranty estimate are never touched, and a lookup that could not run changes nothing. It takes about 15 seconds. An item created from a scanned paper receipt has no such button: delete a wrong timer from the item and add the right one as a Countdown timer. **Timer outcomes are reported (v1.51.144):** when a forwarded email creates an item, the inbox message states each timer's outcome. Since v1.51.185 the warranty and return research runs only when what was bought is a durable good (a machine, tool, appliance, electronics, furniture, a vehicle, a plant); a ticket, a course or book, a download, a service, a subscription or supplies gets no timer, and the inbox line says so. **Since v1.52.16** a thing its maker warrants counts as a durable good even though it wears (tires, batteries, mattresses, footwear), and the receipt's own words outrank the reading: if the receipt names a warranty, protection plan or guarantee for the item, the research runs and the record is a product, not a plain Item. **Seller's promises (v1.52.17):** every dated promise the seller makes on the receipt — a nursery's survival or tree-and-shrub guarantee, a breeder's health guarantee, a contractor's workmanship guarantee, a protection or service plan sold with the item, a dealer's own 90-day warranty — becomes its own countdown timer named as the receipt names it, dated from the purchase, carrying the receipt's words and any stated conditions. It is read from the receipt only (text of a forwarded email, pages of a scan) — never researched on the web, never guessed from the item's name — and it works for every kind of purchase, plants and animals included. It shows on the Activity tab, the Due list, reminder emails and as its own row on the Coverage tab (" · seller's promise"); for a forwarded email the inbox note lists it with the other timers. If a guarantee was not picked up, add it from the item as a Countdown timer — research always finds *something*, so it is not asked — the warranty and return-window timers that were set (with end dates), an already-expired warranty, or an honest "couldn't verify" with the next step (add it from the item's Coverage tab). A lookup that could not run is tried once more; one that ran and found nothing is reported as it is (v1.52.50). Items whose only evidence is a receipt (no photo) show a "Receipt on file" tile instead of "Add Photo" — tap it to open the file. An emailed receipt's view has a Print / Save as PDF button that opens a clean copy in its own tab for the browser's print dialog. **Receipt honesty and freshness (v1.51.139):** an emailed receipt's price is what you actually paid — discounts are subtracted, and the printed total minus tax always wins over a misread subtotal. The receipt is kept once, as a PDF with a first-page preview; every uploaded PDF now shows its first page as its thumbnail (gallery, log evidence, item header). Returning to the Qadvo tab checks for new items immediately — forward an email, switch back, and it appears without refreshing. **Email imports and log re-scans (v1.51.138):** forwarding an emailed receipt with SEVERAL machines on one order creates one record per machine (price split, tax pro-rated, own warranty/return lookups), each with the receipt attached as a PDF and the order facts in the notes; add a photo of each serial-number label so coverage checks can identify them. A LOG's attached files can be RE-SCANNED by AI — from the file viewer or inside Edit Log — and the read merges into that log (fills missing cost, tax, vendor, meter readings; adds detail; never deletes; never touches the item's fields); with several files you pick which to read together — picked files show the orange border and check (orange = AI throughout). The coverage row says "No warranty on record" when nothing is known, never a placeholder claim. Condition is one card: when purchased (New…Poor, the acquisition fact) and now (Excellent…Damaged, drives the Condition report). **Years of records (v1.51.137):** the $/year window starts at the purchase date when the first log is within 12 months of it (records that begin near purchase are complete from purchase); otherwise at the first log (silence before a late-starting record is missing logs, not zero spend); floor one year. Hover the column for the rule and which date a row uses. **Purchase-document tax (v1.51.137):** scanning a bill of sale, buyer's order, or quotation onto an item fills Price (pre-tax), Tax, and Tax rate — the same convention as logs — and never overwrites numbers you already entered. **Economics tab, honest edition (v1.51.136):** the default comparison is $ / YEAR (lifetime operating ÷ years of records) with a small count of the priced logs behind it, plus $ / UNIT and a 12-MONTH TREND (▲ up / ▼ down / → flat vs the prior 12 months; STALE when nothing was logged in the last year — silence is not improvement). % of purchase is in the column picker (it punishes cheap, long-lived machines). A metric shows only when its inputs cover the row: Value shows a SET chip (one tap opens that item's $ Value screen over the Overview) when no current value is recorded — the purchase price is never shown as a value; group values say "$7,500 · 1 of 17" when sparse; $/unit on a group needs every costed member metering in one unit; the "operating is N% of recorded value" figure appears only once valued items carry half the fleet's purchase basis. $/year and $/unit are RATES (headers marked ÷) — member rows never add up to a group row. Nesting works like the Items tab (chevrons; child records such as a flatbed indent under their truck and roll into the truck's Improvements $, kept out of operating). The header dollar badge shows BASIS when it is mostly purchase prices standing in for values. **Overview → Economics (v1.51.135):** the Overview page has four tabs — Items, Logs, Maintenance, and **Economics**. Economics lists every item with purchase price, value, lifetime operating $ (consumables + maintenance + records; improvements shown apart), % of purchase (over 100% = it has cost more to run than to buy), this year, $/unit and usage where a meter is logged, priced logs and last log; sortable by any column (default: most expensive to run first), searchable, column picker; group rows roll up their members plus the shared logs on the group and on every group inside it (since v1.52.36 the Overview, Q and By month count a subgroup's own logs, as the group's Economics tab always did); fleet totals on top. On every Overview tab you can drag a column header to reorder and drag its edge to resize — both stick per tab; Reset restores defaults. Ask Q answers portfolio cost questions ("which truck costs me the most to own?", "are any trucks due for replacement?") from the same per-item figures. **One set of numbers:** Q's cost answers — a fleet question, or "cost per mile" with one item open — come from the same calculation as the Economics tab and the Overview, so Q states the tab's figures instead of working them out from the logs. Amounts are before tax, as the tab shows them (sales tax on a purchase or a repair is not in these figures), and refunds are already subtracted. Money counts whatever a log is called: a reading or inspection log that carries a cost (a billed shop ticket that an import filed as a reading) counts as maintenance on the Economics tab, the Overview, By month, Q and the Certified History — before v1.52.36 those costs were left out — and the invoice importer no longer files a billed paper as a reading, however prominently it prints the meter. A machine logged per job (2 hours, 3.5 hours, 2 hours) has its hours ADDED UP; a running meter (odometer, hour meter) counts from the first reading to the latest. When a meter reading doesn't fit the rest of an item's log — a truck at 231,871 miles in March and 127,965 in July — Q names that reading as a likely typo. The usage follows the meter's longest run of readings that never goes down, so a reading that doesn't fit is LEFT OUT — on the Economics tab (which names it at the top, in the amber **Check these records** card, as a reading "out of sequence with the meter's other readings — not used for miles/hours" (a meter only counts up)), the Overview and Q alike; correcting that log makes it count again. **Marked wherever it shows (v1.52.53):** the item's Performance Monitor says "N reading(s) not used" on that meter, with an amber dot on its sparkline; the meter's chart (tap the meter) puts an amber dot on each one, names them under the chart and marks them "out of sequence with the other readings — not used for miles" in its entries; and the log the reading was typed on shows "check reading" in the lifecycle log (open it: the reading's chip reads "— out of sequence with the other Odometer readings, not used for miles"). All of them use the one rule — none decides on its own. **What "not used" means:** only the READING is set aside — the miles or hours are counted without it; the cost on that log still counts in every figure. Each one is named with the counted reading it clashes with ("48,210 on May 2, 2025, 3 days earlier, is higher"), because the reading left out isn't always the wrong one: the rule keeps the side with more readings, so a ticket with the right truck and the right miles can be the one set aside when a neighbouring ticket is off or belongs to another item. Tell the owner to check the papers on BOTH logs: fix whichever reading is wrong (never change a reading to something its paper doesn't say), or move a log that is another item's to that item (the log's move button) — its cost goes with it. Leaving a reading out doesn't change a machine's lifetime miles (they run from its first counted reading to its last); it can move miles between seasons, so a season's cost per mile can shift while every dollar stays where it was paid. **Cost per mile or hour with no fuel logged (v1.52.53):** the Operating Costs section shows logged cost over logged use whenever the logs give it — the same figure as the top of the tab, the Overview, Compare and Q — marked "logged" and saying what it leaves out ("no fuel logged", or the fuel and supplies it includes); "measured" means fill-ups cover most of the miles or hours. Before v1.52.53 it said "Not measurable yet" without fuel fill-ups, while the rest of the app showed the figure. Scanned and imported invoices now keep a meter reading's unit: an hour meter read off an equipment invoice is saved as an "Hour Meter" in hours (it used to become "Odometer, miles"), the review screen shows the unit next to the reading and lets you change it, and a reading lands on the meter the item already tracks. An odometer or hour meter is always read as a running meter — its readings are never added up, even when a few of them run backwards (a reading copied from another truck); only a per-job or per-run log (hours per job, a shoe's distance per run) adds up. **Readings from another item:** when the same meter reading appears on two of your items within a month (a shop ticket filed under both), it counts only on the item whose log it fits — on the other it is LEFT OUT of usage everywhere and listed on its Economics tab as "Left out as another item's readings" with where it came from; the log's COST still counts. A reading that fits both logs stays and is listed as "Also on another item — check the paper". A part filed under a vehicle (a tire under a car) that logs the vehicle's odometer shares that meter: its readings are never called another item's (v1.52.55). **Logs that look like another item's** (an amber card at the top of an item's Economics tab, only when there are some): a job on two items — the same date and cost plus the same invoice number, meter reading or scanned file — and another item's reading on a job only this one has. **The amber "?" on the Economics tab (v1.52.56):** an item's Economics tab carries an amber count with a question mark ("3?") when it has readings to check — the readings Check these records lists (ones that don't fit the others, another item's, or on another item too), each counted once; a group's Economics tab, and the Overview's, shows "?" when any item under it has some, and lists those items at the top of the Economics view (tap one to open it). Pointing at the mark says what it means. The "are records missing?" question and whole-job copies show on the tab but aren't in the count, and neither are paper differences. Each shows both logs with their tickets and what the meter readings suggest (or that they don't settle it); the owner or an admin can **Remove it from this machine** / **from the other machine** (the other keeps the job), **Clear the reading** (the cost stays) or **Move the log** to the machine it belongs to. Every action takes a second tap, and nothing is deleted outright: a removed copy is kept whole in the note left on the timeline, a cleared reading is named in its note, and a move leaves a note on both machines. Those notes sit at the top of the machine's Activity timeline (dated the day of the fix, marked System); a removed copy or a cleared reading can be undone from its note with **Put back** (a second tap confirms; owner or admin) — the log, or the reading, returns as it was and a new note records who put it back. Fix notes are locked: they can't be edited or deleted, so what they hold is never lost. The invoice importer now checks every scanned ticket against the machine it is going onto — its VIN, plate, model year, the machine number written on it, and whether its meter reading can fit that machine's log — and says so on the row ("This ticket may not be Truck 18's: its reading 216,140 is above this machine's 178,520"); when the paper names exactly one other machine it offers **File it under …**, and each ticket can go to a different machine in one import. A machine number the shop wrote can be wrong — the meter reading is the stronger evidence, and the owner decides. **Compare view** ("Cost per mile by season" / "Cost per hour by season"): on a group's Economics tab and on Overview → Economics, a **Table | Compare** switch shows every machine on one chart — across, how much it was used in a working season (hours or miles); up, what each hour or mile cost to operate; both on log scales so a mower and a loader fit together. Each dot is an item's last full season, labelled with its short name (a hollow dot is the season in progress; a faint, small dot rests on few priced logs; a dashed ring means nothing has been priced since that season; a grey square is an item not in active service — retired, sold, traded in, scrapped, lost, died — labelled with its status and included by default for history, with one "Include inactive" box to hide them all; an amber diamond marks readings left out or to check — likely typos, or another machine's readings). Below the chart, **What stands out** says it in words: costing the most and the least per mile or hour, machines whose cost per unit rose three priced seasons in a row, little use, and nothing priced since. **Its usual:** a machine with at least two complete measured seasons has a usual — its usual SEASON total (operating cost over those seasons ÷ their number), its usual use a season and its usual cost per unit — drawn as a soft DIAGONAL: that machine's own line of equal season cost, across only the range of use it has actually seen, with a small mark at its usual use. A dot above its line cost more than its usual season, below it less; left of the mark it was used less, right of it more; a thin arrow runs from the mark to this season's dot. Every season is judged on BOTH counts, never cost per unit alone (a machine used less costs more per mile even when it costs less in total): **costlier than usual** (more in total and per unit), **cheaper than usual**, **used less, not costlier**, **used more than usual**, or **about usual** (both within 20%) — on the dot's line, its card (with the three numbers: season total, use, cost per unit, each as × its usual) and Ask Q. Shown for the tapped machine, or for all with "Show usual costs"; point at a line to see whose it is. "What stands out" lists machines **costlier than their usual** (1.5× its usual season or more, and more per unit, with 3+ measured seasons) — a normally cheap truck having a bad year. **Use changed:** two kinds, told apart. A **step** — its job changed: use per season halved (or doubled) and LEVELLED there, at least two measured seasons on each side, e.g. a front-line truck that became part-time. "What stands out" lists it and the card asks **Was it reassigned?**; the owner or an admin answers: **Yes — its job changed** makes its usual measured from that season on (older seasons are read against its usual at the time), **No — keep its whole history** leaves it; the answer is saved as a locked note on the timeline. A **slide** — use fell (or rose) season after season — is only described ("use fell season after season, from about 10,300 mi in 2021 to 1,800 in 2023"), never asked about, and its usual stays its whole history. When use fell by half or more while the season cost didn't follow it down, the card says so ("its cost didn't follow its use"). A machine no longer in service is never asked: its card looks back instead ("its use fell about 83% over 3 seasons while it kept costing about $5,000 a season — sold May 27, 2026"). When an active machine shows that pattern and one of the owner's retired or sold machines showed it too, the card says **One of yours went the same way** — the owner's own past decision, stated as a fact from the records, not a recommendation. **Records look thin:** a season with at least its usual use but well under a quarter of its usual cost is faded on the chart (amber dashed ring), listed in "What stands out" and asked on the card, one question at a time — **Are records missing?** (since v1.52.53 the machine's own Economics tab asks it too, in **Check these records** at the top, with the same answers saved the same way) Low cost alone never removes a season: a cheap year looks the same, so it stays in its usual until someone answers. **Yes — records are missing** takes it out of its usual (saved as a locked note; Q then calls that season under-recorded); **No — it was a cheap year** clears the flag. The season line reads "Seasons run January – December" with no qualifier when the calendar-year default is in effect. The usual's two limits, always stated with it: it **lags** a steady climb (getting worse pulls the usual up, so the gap understates the trend) and it adds up **different years' dollars** with no inflation adjustment. **All seasons** (in the season menu) puts each machine at its whole measured history instead — its usual use a season across, its cost per unit over those seasons up — to compare whole working lives, sold machines included, without one big repair deciding it; it hides trends, and a machine whose use changed has one dot describing neither job. **Fuel:** the chart's header says whether fuel is in the numbers — "Fuel isn't logged for these machines — operating here is maintenance, repairs and renewals, not all-in", or which machines have fuel logged and how much; a machine's card says "(no fuel logged)" or how much of a season was fuel & consumables. Once fuel-ups are logged every dot moves up, and "What stands out" marks a rise that starts where fuel first appears as fuel arriving in the records, not the machine getting worse. A machine not in active service shows its status — Sold, Traded in, Retired … with the date — right after its name on the line above the chart. The verdict's words are on that line and the card, not inside the chart. Tap a dot or a name to see its **trail** across seasons (a trail runs from older seasons to newer: its segments fade from old to new, and the tapped machine's carry small arrowheads pointing forward in time) and its numbers (tap empty space to clear; the chart's scale never moves while you explore); "Show all trails" draws every machine's. The **diagonal lines are equal season cost** ($1K, $5K a season…): left and high on a low diagonal means idle (little spent, little used), left and high on a high diagonal means expensive (a lot spent on little use); a machine climbing across diagonals season after season is the one to look at. Hours and miles can't share an axis, so chips switch between them. A season's cost per mile or hour is stated only when the meter readings span most of it (60% — the Economics tab's own bar for a measured cost per unit): a season with costs but no meter reading has use that is unknown, not zero, and a season whose readings stop early would divide a whole season of costs by a few weeks of use. Those machines are named right under the chart with their last reading date ("Costs logged in 2024, but no meter reading: Truck 17 · sold (last reading Jan 6, 2023)"), and Q says the same instead of dividing — logging a reading fixes it. Every machine that isn't plotted is listed under "Not on the chart" with the reason (no use logged, use but no costs, measured in the other unit, no longer in service). Operating includes renewals like registration or insurance when they're logged to the machine, before tax. **Working seasons** are the 12 months from a start month, because a season means different things by trade and by place (a landscaper's year, a crop year, a winter contract). The chart states it at the top — "A season here runs January through December" — and every season it shows carries its dates; change the month right there (the account's, or tick "Just for this group"; a group's setting reaches its subgroups and items). The chart shows the same numbers as the Economics tab and Q; it can't see operator skill, job conditions or how hard each season's work was. Only one is active at a time (it turns orange): **All** (everything, incl. sub-items), **Main** (top-level items only — the cleanest scan), **Due** (overdue/upcoming maintenance), **Shared** (shared with you or by you), **Archive** (archived items only). #### Coverage & insurance **Coverage adequacy (v1.51.87):** ask Q "do I have enough coverage?" or "how does my belongings value compare against my policy?" on a room or property — Q rolls up the documented contents from room inventories (stating the owner-valued vs AI-estimated split), flags categories over typical policy sublimits, and compares against the policy's personal-property / Coverage C limit when recorded. If the policy record lacks its Coverage C limit, Q says so and suggests logging it (or re-scanning the declarations page) so the gap check has both sides. Open **Portfolio + Coverage** (top-right of the header) to see every policy; tap a policy for its detail. **Link a policy** to a property, vehicle, or group by adding it as a child of that parent — open the parent → Links/Contents → add the policy, or upload a policy document directly into the parent. Q then walks up the tree, so when you ask about an item nested under that parent it surfaces the relevant coverage automatically (e.g. a washer inside a house finds the homeowner's policy). Use **Annual Premium** for what you pay each period and **Coverage Amount** for the protected value. **What it covers (v1.52.25):** every insurance policy, warranty and protection plan shows a **What it covers** card on its Info tab, under the carrier and term. By default a policy covers everything under where it is filed, and the card says so with a count of the items owned today (tap the count to see them). When that isn't true, tap **Change** and check exactly the groups and items it covers, for example a fleet auto policy filed under one group of vehicles that also covers another group, or an equipment floater filed at the top of the account that covers only the equipment groups. A group means whatever is in it on the day asked about (a truck bought next month is covered once it's in the group; a sold one drops out), and an item brings what's attached under it. Once set, the policy covers exactly those: it narrows as well as extends, so a policy set to cover the equipment no longer shows up for the vehicles even though it's filed above them. Clearing every box puts it back to where it is filed. Only the account's owner or a workspace admin can change it, because it reaches across the whole account; each change leaves a note on the policy's timeline. The other direction is on every item's Info tab: **Covered by** lists each policy that covers it by the same rule, with what that policy covers; tap one to open it. **Coverage check (v1.52.26):** Portfolio + Coverage, and every group's Info tab, show a **Coverage check**: the things the records say to ask your agent about. It lists each vehicle no policy on file covers (with the last date anything was logged on it), other items no policy on file covers by the group they're in (with their recorded value), any item two policies cover (you may be paying twice), policies whose term has ended (update their dates if renewed, archive them if not; since v1.52.27, when an ended policy was the only one on file for a vehicle or item, the check says so on that policy — " ended on — the only policy on file for " — instead of saying no policy on file covers it: if it was renewed, update its Renewal date on the policy — a logged payment or a new scan doesn't change the dates on file; if not, ask your agent), and policies that cover nothing you own today (filed under something sold, or at the top with nothing set). It uses the same rule as What it covers. **Retired items are parked (v1.52.53):** an item marked Retired (out of service but still yours), or anything kept under one, isn't listed as having no policy on file — a parked machine often carries no policy on purpose. Sold, traded, scrapped or lost items were already left out. Paperwork doesn't count as a second policy: records that share a policy number are one policy (an insurance ID card carries its policy's number), and a record with no number, premium or limits (a lender's insurance selection form) isn't coverage. Warranties aren't insurance and aren't counted. It says "no policy on file covers it", never "uninsured", because a policy that isn't in Qadvo can't be seen. It stays hidden for anyone who keeps no insurance in Qadvo. Ask Q "any gaps in my insurance?" or "is this truck insured?" and it answers from the same check. **A policy's schedule (v1.52.27):** a policy that lists what it covers item by item — the covered autos with their VINs, the scheduled equipment with its serials and values — keeps that list on the record, and the check matches each listed item to your records by VIN or serial. It then also raises: a machine the policy still lists that your records say you sold (you may be paying to insure it), listed items that match no record (add the serial to the record if you still have it; ask your agent to take it off if you don't), a VIN or serial that differs by one character from your record's (check which is right — a claim can stall on it), and equipment attached to an item (a flatbed on a truck) that the policy doesn't declare when it declares extras on other items. New scans read the schedule in the same pass. A policy scanned before that shows **Read its schedule** on its What it covers card — one AI read of what's already on file, no re-upload; the card then lists every scheduled item with the record it matches or why it doesn't. Items the policy lists without a serial can't be matched and are never called missing. Ask Q's coverage answers ("is this truck insured?", "what covers this machine?") and the Insurance Inventory report use the same rule, and each policy there states what it covers. The owner's choice is stored with the policy, not by moving records: where a record is filed stays where you find it. **A policy's term today (v1.52.27):** every insurance policy's Info tab shows its **Term today** — in force through a date, ended on a date, or (for a policy that renews automatically) "its renewal isn't on file". Paying the premium keeps it current: **log each term's premium on the policy** — the amount and the day it was paid — and a premium logged from two months before a renewal date to one month after carries the policy into that term. A forwarded bill or payment email reaches a policy only by hand today: put "log" at the front of the subject, pick the policy on its Inbox card, and fill in the amount if it's blank (an email saying a payment is SCHEDULED isn't a receipt, so its amount isn't filled in, and one payment for several policies — an auto and a homeowners policy on one bill — belongs to all of them, which Qadvo can't yet record) (a premium near a later renewal carries it through the terms before). For that, Qadvo needs to know how often it renews: it reads it from the dates on file (effective to renewal), or set **Renews every … months** on the Term today card, with **Automatically** ticked for a policy that renews on its own (six-month auto policies usually do). An automatically renewing policy past its date with no premium logged isn't called ended — the coverage check, the item's Covered by row and Ask Q say its renewal isn't on file, and don't count it as coverage until a premium or new dates are on file. Scanning a policy's NEW declarations page onto it also renews it: a scanned term that ends later than the one on file replaces the dates and premium, and the policy's timeline keeps the old term; a scan of the same term still only fills blanks. Every item's **Covered by** row now says whether each policy is in force today (green) or ended / renewal not on file (amber), the Insurance Inventory report prints each policy's term on its date, and Ask Q states a policy's term instead of assuming it's active. **What it covers** now starts from what the policy covers today: with nothing set, the item or group it's filed under is already picked, so picking a trailer for a car's policy ADDS the trailer; removing the item it's filed on narrows it, and the card warns when a policy is set to leave out the item it's filed on. **VIN check digit:** a 17-character VIN's 9th character is a check digit (every North American vehicle and trailer since 1981), so a VIN with one misread character fails it — a vehicle's Info tab says so under its VIN ("check it against the plate or the registration"), and when a policy's schedule lists a VIN one character from the record's, the coverage check says which one the check digit fits. Since v1.52.55 a purchase paper's 17-character equipment PIN fills the machine's serial number, never its VIN, and a new item made from it is equipment, not a vehicle. A machine's 17-character PIN (a skid steer's, a mini skid's) is a different form: its 9th character is a letter and it has no VIN check digit, so it's never judged by one — before v1.52.38 every equipment PIN showed "fails its check digit" on a correct serial. #### Portfolio value The **Portfolio + Coverage** tile in the header shows your total asset value and total insurance coverage. Tap it for **Portfolio Overview** — a breakdown by type (real estate, vehicles, equipment, animals/people), debt and equity totals, and your active policies. #### Quantity (one record, many units) An item can represent multiple identical units — 5 of the same tree, 40 fence posts — via **Qty** (set in Edit Details or when adding manually; receipt scans pick up a quantity only when the paper explicitly states one, never guessed). **Money rule:** the **purchase cost is the LINE total** (what was paid for ALL units — the cost field relabels itself "Cost (all 5)" and shows the per-each price as you type), while **Current Value and Replacement Cost are PER-EACH** ("$200 each"). Everywhere totals appear — the item header ("$1,000 (×5)"), the $ Value screen, dashboard, Portfolio Overview, group contents, Economics, reports, and exports — the app multiplies per-each value by quantity automatically. The Overview → Assets table has a Qty column (on by default). Setting Qty back to 1 clears it. If a user asks why a value shows "(each)" or "×5," this is why. #### Hiding a value, or opting a document into one Each item's **$ Value** screen has a per-item control over whether it carries a monetary value — the owner decides, never the category. On a normal item, **"Value not applicable"** hides the dollar value and the "Set value" nudge from its card and header everywhere and drops it from totals (for a body, a freebie, anything where a price is just noise). On a **document** (a title, deed, certificate, or "Other" QDoc), the flip side appears — **"This document has a monetary value"** opts it IN, so a rare book, a collectible certificate, or an appraised deed can carry a value like any asset (with the AI sell/market-comp lookup) and count in totals. Insurance/warranty documents keep showing their coverage limit instead. Both toggles are reversible anytime from the same Value screen. #### Overview (desktop) — the spreadsheet view The **Overview** button in the header (PC) opens a spreadsheet-style view: **Assets** (sortable, column picker), **Logs** (every event across every asset, searchable/groupable), and **Maintenance** (every interval and warranty in one list, overdue first — the place to plan service work). Click any row to jump to that asset. The Assets tab opens **nested by default**, as a table of contents — top-level items and their direct children shown, deeper groups closed (v1.51.184); one **Expand all / Collapse all** pill beside Show nesting opens or closes every group, the chevrons work per group, and the session remembers what you opened. Groups cluster their children with **collapse chevrons**, and each group row shows its **count (Qty) and subtree total value** in indigo — everything inside, counting item records AND ledger lines (a container whose only contents are ledger lines, like a Shop, is a collapsible parent too). **Column widths are draggable** (grab a header's right edge) on all three tabs; the header stays frozen while you scroll, ledger contents appear as rows alongside records, and Qty always shows a number (group rows show how many items are inside). "Show nesting" can be toggled off for a flat list. **The Overview remembers where you were** for the browser session: leave to look at an item and come back, and you land on the same tab with the same search, sort, filters, and collapsed groups; each tab's **Reset** button clears them. Column widths and order persist across visits. **Search finds logs by amount or invoice number (v1.52.37):** type an amount — "412.80", "$412.80", the total with tax, or a whole dollar like "412" — or an invoice / ticket number ("N2214", "n-2214": case, spaces and dashes don't matter) into the home screen's search, and the record holding that log shows up; its card says which log matched ("· log: Apr 18, 2025 · $437.57 · #N2214"), and opening it lands on the Activity tab. A credit is found by its figure without the minus. A number of one or two digits is still a name or unit number ("12" finds Truck 12), so an amount needs three or more digits, a decimal point, or a $. The Overview's **Logs** tab search matches amounts and invoice numbers the same way. #### Where you left off — refresh, the Back button, and links to an item The address bar tracks where you are: the open item and its tab, a full-page view (Overview, Inbox, Settings, Portfolio), or a home filter such as Due. **Refreshing the page brings you back to the same place** instead of the home screen. **The browser Back button (or a phone's back gesture) returns to the previous item or view** instead of leaving the app. **Copying the address gives a link straight to that item and tab** for anyone who already has access to it. If the session had expired, signing in returns you to the same place. Overview filters and sort are not in the address; they are remembered for the session as described above. **Portfolio value includes ledger contents:** the header total, Portfolio Overview, By Type, and By Group all count contents-ledger lines (value × count) along with item records — counted small stuff is owned value, so Qadvo's totals reconcile with a business's own spreadsheet. #### Economics / cost of ownership The **Economics** tab on an asset tracks total cost of ownership — purchase price, maintenance/repair costs, depreciation, and cost-per-hour/mile — useful for repair-vs-replace decisions. To compare spending ACROSS everything (e.g. "what's costing me the most to own?"), just ask Q — it analyzes your whole portfolio, no per-asset setup needed. **By month (v1.52.36):** an item's Economics tab shows **By month** — what it cost each calendar month, in bars by kind of cost (maintenance, repairs & records, fuel & supplies; a log is fuel & supplies when it carries a fuel amount — a metric named Fuel, Diesel or Gas, or Gallons on a fill-up — or a supply it tracks, or when it has no kind of its own and reads like a fill-up; a log you typed Maintenance, Record or Improvement keeps that kind whatever its notes say, and a plain "Gallons" there is the oil or hydraulic fluid the job took, not fuel, so it never counts toward fuel per hour — name a fluid for what it is, v1.52.37), with "Running about $X a month" (the average of the last three full months, while it's in service) or else the average of up to 12 months, plus last month, this month so far and the same month a year before; 12 months, 24 months or All (rolling windows that end with this month, not calendar years; the headline figures don't change with them). The chart opens on the shortest of those that holds a cost — records that end last year (a machine's old shop tickets) open on 24 months or All — and with nothing logged in the last 12 months the headline says so and names the month of the last cost. A group's Economics and Overview → Economics have a third view beside Table and Compare, **By month**, split by what is directly under where you are: on the Overview, each top-level group as one part and each item that isn't in a group on its own; on a group, each subgroup as one part, its items, and the group's own shared logs (an item kept under another item — a furnace under a house — is its own part, its money being its own). Tap **›** beside a group to look inside it: the chart, its headline and its figures become that group's, and the trail above it (Everything › Farm › Trucks) steps back out; a place that holds only one group (a farm kept under one group, say) opens inside it. Every level adds up — a group's part on the level above is that group's own total. Parts with costs outside the months shown, or folded into "Everything else", are listed under the chart with their names and **›**, so every one can still be opened or looked inside. Parts with nothing logged in them are listed ("No costs logged in: …"); sold or retired members aren't counted; tap a name to open it. Tap the **colored box** in front of a name to hide that part for a look — e.g. what a business costs to run without one big subscription: the bars, the month totals and "Running about $X a month" are then figured without it, the chart says "Not shown: …" with **Show all**, and the box turns hollow. Nothing is saved; the chart opens with everything shown, and it changes no record, total elsewhere, or Q's figures. The Overview's Economics table gains monthly columns — **Last month $** and **Avg / month $** (shown) and **This month so far $** and **Same month last year $** (in the column picker). The months come from the same economics code as every other figure, so Q states them: "what did I spend in August?", "the monthly cost of …", "the most expensive month", "this month against the same month last year" (the fair comparison for a seasonal business). Amounts are before tax with refunds subtracted; improvements and purchases aren't monthly costs; a month with nothing priced logged says "no logs" — missing logs or a quiet month, not a measured $0; months before a record's first entry aren't shown; a yearly charge (insurance, registration, a domain) sits in the month it was paid — the "running about" figure doesn't spread it yet. "Running about $X a month" is never stated over three months with nothing logged (the chart then gives the 12-month average, or nothing) — silence isn't $0. #### Activity tab Two parts: a **Performance Monitor** (metric sparklines, interval progress bars) up top, and the **Timeline** (chronological log of all events) below — the day-to-day view per asset. The Timeline's condensed view shows each entry's amount as paid, to the cent (cost plus tax) (v1.52.31). **Move a log (v1.52.37):** open a log and tap its **⇄** (between the pencil and the trash; owner or admin) to move it, as it is — date, amount, files, readings — to another record of the account, picked from a searchable list with the records near it first; a note on both records says where it went, and a forwarded receipt filed as that log follows it (its Undo still works). It's how a paper parked on a group goes to the right machine once you know which. **Units on tracked metrics:** a tracked metric can be a plain number with NO unit (a confidence level, a score, a count) — leave the unit box empty in +Log and it stays empty everywhere; it never borrows a unit from another metric in the same log. Each metric keeps ONE unit: when you log by voice (Speak) or through Ask Q, the value lands in the unit that metric is already kept in — say "45 minutes" on a metric kept in hours and it logs 0.75 hours; say it on one kept in minutes and it logs 45 — so the sparkline stays on one scale. If what you said can't be converted to the metric's unit (pounds vs gallons), the log is saved as spoken and flagged in your Inbox to review. **Timers are different:** a Service Interval or Countdown always needs a unit, because it means "every N days / miles / hours". The prebuilt interval collections (Truck/Car, Tractor/Equipment, Zero-Turn Mower, Home: HVAC/Seasonal/Appliances) and the quick-fill chips (Oil Change, Hydraulic Fluid…) are offered by WHAT THE RECORD SAYS THE ITEM IS — its name, maker, model and category words — not by whether it was filed as Equipment or Item (that line is the owner's habit: furnaces and dishwashers are usually filed as Equipment). A furnace gets Home: HVAC & Safety and a Filter Change chip whichever way it is filed; a fridge gets Home: Appliances; a skid loader gets Tractor/Equipment; the filing type is only the fallback when the name says nothing. A person, plant, animal, policy or group sees none of them (v1.52.5). The Alert slider on a timer goes to 100% = **Alert when due** — the reminder fires on the due date / at the limit, not before (it was capped at 95%). **After every save** (+Log, Ask Q's "Yes, log it", a fuel or service-card review) the confirmation names the item's most urgent timer — "Next up: Tire Rotation due in 2,903 miles (71% used)" — ranked by the share left, so miles and days compare, in the item page's own numbers (the newest reading by date, never a log entered from old records); an ended warranty is history and never shows there (v1.52.32). To DELETE a timer: click it in the Performance Monitor, tap the trash icon beside Edit Timer, then tap it again within 3 seconds (v1.52.6) — or from the Countdowns & Intervals list, or the Edit Timer wizard. Deleting removes the timer and its settings; the logs that reset it stay on the timeline. Setting up ANY timer starts with one choice, **Count by: Calendar or A meter** — the same in the timer wizard (new from the Performance Monitor's **+**, or edit by clicking a timer) and in a +Log metric row set to Interval or Countdown. **Calendar** (days, weeks, months, years) needs nothing else and is how to set a plain reminder like "once a week" on anything, including a person or a policy with no meter. **A meter** (miles, hours, cycles…) then asks which tracked reading to count against. In a +Log row the interval reads as a sentence — "Oil Change, every 10,000 miles" — because that number is HOW OFTEN, not a reading; the reading goes in the meter's own row (Odometer). **Track against** shows the reading THIS log counts from: the log's own meter row, else the meter's latest reading only when it was taken on or before the log's date. A service log dated before the meter's latest reading — one entered from old records — asks for the reading at that service instead of borrowing a later one, which would start the timer from the wrong mileage (v1.52.31). A new timer starts on the calendar unless the item actually tracks a meter. Hours is a meter (run hours), not calendar time. **Log reminders (v1.52.3):** to track a reading on a schedule ("log the tire pressure every 7 days"), add it as a **Track** metric in +Log and tick **Remind me to log this** — every N days/weeks/months. That creates a calendar timer named "Log " that nags you, and **logging the metric restarts it automatically** — by +Log, by voice, or through Ask Q — with no reset chip. Clicking that timer's +Log opens on the metric itself (Performance type, value box empty, unit as you keep it). In the timer wizard the same thing is the **Reminds me to log** picker on a calendar interval. Any timer's +Log now opens on the log type its previous resets used (a service timer → Maintenance, a log reminder → Performance). To track a plain number, add it as a Track metric in +Log, not through the timer wizard — and don't give a tracker the same name as a timer on the same item, or logging that name resets the timer instead of recording a reading. #### Links / nesting (parents & children) **Order of items inside a group (v1.51.123):** the Contents list and a card's expanded children use one definition of "most recent" — the same as the home screen's Main/All: when something was last logged, rolled up through the whole tree, so a group with active members rises. The Contents tab has a **Recent / A–Z** toggle next to the Inventory heading; your choice is remembered on that device. The **Links** tab shows items nested under an asset — parts in a machine, appliances in a home, policies on a property. Move items between parents, add children, or upload documents directly to a parent. #### Info / specs The **Info** tab shows specifications (manufacturer, model, serial number) and coverage details. **Every fact on the Info tab is tap-to-edit where it's shown (v1.52.30)** — tap the value (a small pencil marks it), type, press Enter (Esc cancels): make, model, year, trim, VIN or serial, license plate, asset and label numbers, location, the **Purchase** card (purchased from, purchase date, paid with, miles or hours at purchase), price, tax paid, current value and amount owed. Changing the price recomputes the tax from the rate on file, and changing the tax recomputes the rate; a vehicle's VIN and its serial are saved together. The **Edit** (pencil) window stays for what isn't a single fact — the type (and so the Category), the name, where it lives, photos, archive and delete, and editing everything at once. So when a scan's "Use the scanned one" notice names a purchase fact, that fact is on the Info tab, where it can also be corrected by hand. **Q reads whole records everywhere (v1.51.133):** every document Q consults — the open item, the policies and contracts around it, every item in a count or search — reaches Q as its complete record (notes, transcript, identifiers, dates, terms, every scanned field), trimmed only for size with a visible note, never by dropping fields. **Q reads the whole document (v1.51.132):** coverage questions asked of Q see everything a policy or contract QDoc says — notes, the covered item's serial/model/year, dates, term, premium, and the transcript — so "what model does this policy cover, is it the machine it's linked to?" is answered from the document. **Multi-page QDocs (v1.51.130):** a policy or service agreement uploaded as several PDFs, or one long PDF, is read as ONE document (up to 12 pages): the reader fills the term and computes the renewal date from it, records mileage limits next to dollar limits, copies contract/VIN numbers character by character, and only reports checkbox options that are actually marked; the full transcript is kept from page one so Q can answer from the fine print. Re-scan (Gallery) works on PDFs too and keeps your own notes — a scan replaces only the lines the SAME kind of paper or photo wrote before (v1.52.30) (a bill of sale's lines, a nameplate's, a plant's): a VIN-sticker or nameplate scan never touches what a bill of sale wrote, and a line is never cut apart at an abbreviation like "Inc." or "No." (before this, one scan could erase another paper's lines and leave fragments behind). Use the **Research** button to have Q auto-fill specs from the web. The **Notes** block shows a small "Updated " line telling you when the notes last changed — and the Activity timeline records each change as a muted, LOCKED system entry ("Notes edited" for your edits, "Notes updated by re-scan" when an AI scan refreshed them; locked because they're provenance, like the creation record). Notes from before this feature show no date until their next edit. **Re-scanning from the gallery**: open the item's photos and tap **"Re-scan photos…"** — with several photos you pick which views the AI reads TOGETHER as one identification (a close-up plus a wide shot beats either alone; up to 6); a single photo re-scans in one tap. Scans fill blanks and refresh machine facts but never overwrite what the record already says — and never drop a difference silently (v1.52.30): when a scan reads something DIFFERENT from the record (a purchase date, a price, a VIN, a model, the miles at purchase), a notice shows both values and **Use the scanned one** takes the scan's with one tap; the notice stays until you act on it or close it. While a scan runs, the button/overlay shows staged status ("Reading photos…" → "Identifying…" → "Updating record…") so the wait reads as progress, and when it finishes a dismissible message summarizes what it actually saved (e.g. "Updated notes, Found model number") — or says "no new details" — so the update is never a mystery. #### Signing in and out (v1.51.187) **Log out** ends the session on that device only — every other device and browser stays signed in (deleting the account signs out everywhere). The app treats the session it holds as "signed in"; a momentary server or provider error never signs you out, and a page that was never signed in shows the landing page, not the login screen. Sessions are checked by Qadvo's own server, not only by the sign-in provider. #### Sharing **Sharing an item vs adding someone to your account — two different things.** **Sharing** (the **Share** button on an item or group, Invite by email) gives one person access to THAT item and everything nested under it, and nothing else of yours. They use their own Qadvo account (free to view and help), see it under their Shared items, and their role — viewer, editor or admin — applies to that tree only. Use it for a customer, a mechanic, a family member: someone who needs one thing. **Adding someone to your account** (**Settings → Account → Team**, a workspace member) lets them work INSIDE your whole account with their own login: an **admin** manages everything except billing and transfers, an **editor** adds items and logs and edits contents, a **viewer** sees everything and changes nothing. They switch between their own things and your account by tapping the name at the top of the home screen, and what they add lands in your records with their name on it. Use it for your office manager or crew: people who work across everything. Neither shares your password — each person signs in as themselves. **Removing someone's access:** open the item or group you shared and tap **Share**. Everyone with access is listed under the invite, each with their role (tap the role to change it — editor to viewer, say — instead of removing them) and **Active**, **Pending** or **Expired**. Tap **Remove access** beside a person to take their access away, or **Withdraw invite** for an invite they haven't accepted (an expired invite shows **Remove**). It takes effect at once: they lose the item and everything nested under it, and their Inbox shows "Share removed" ("Share withdrawn" for an invite they hadn't accepted). Access that comes through a parent is removed on that parent, where the share lives. Nothing they did is lost: their logs stay on the timeline under their name, and anything they added under your shared item is already yours. **A workspace member** (someone working inside your account with their own login): **Settings → Account → Team**, tap **Remove** beside them and **Tap again** to confirm (a pending invite says **Withdraw**); what they added stays in your records. A member can leave on their own from the same page (**Leave**). **A public link:** Share → **Public link** → **Revoke**, then **Revoke permanently** — anyone holding the old link loses it. On a group's **Activity** tab the **Public view** pill previews what the group's public link shows — one combined, read-only log of everything in the group. It needs a public link first (Share → Public link); without one it says so. **Sharing a parent shares everything under it** — share a property or group and every item nested inside (at any depth) comes with it; access flows only DOWN the tree. **Anything added under a shared parent belongs to the parent's OWNER**, not the person who added it — whether added by hand or by photo scan (v1.51.79) — so the owner always sees their whole tree; the timeline's "Asset Record Created" entry records who actually added it, and adding requires editor or admin access on that parent. **Spotting shared items (v1.51.79):** dashboard cards, their contents rows, and the Overview tables all show a green users icon — arrow pointing IN = shared with you, arrow pointing OUT (↗) = you're sharing it — including items shared only because a parent is shared. On an item's page the **Share button itself** turns green with a small ↗ when the item is reachable through a shared parent (hover names the parent). **Inherited access in the Share screen (v1.51.79):** open Share on an item inside a shared tree and a read-only "Inherited Access" list shows who can reach it, their role, and via which parent — manage their access on that parent, where the share lives. Recipients see "Your access comes through " on inherited items, and leaving a share warns it removes access to that whole shared tree, not just the one item. **Moving shared items (v1.51.80):** a person you've shared with (editor/admin) may move an item between groups belonging to the SAME owner — reorganizing your tree for you is allowed, and ownership doesn't change — but can never move an item OUT of the owner's account: detaching it to top level or moving it under a different owner's tree is owner-only and blocked. Permanent handoffs between accounts use **Transfer Ownership**, not a move. **Privacy at transfer (v1.51.96):** when ownership transfers, every email address in the record's inherited history is masked (j•••@y•••.com) — previous owners and their sharees never travel in readable form to future owners. The new owner sees only their own transfer counterparty in full; the previous owner's archive copy keeps its complete history. **Every save ends with a next (v1.51.119):** after logging from Ask Q, a magnet service card, or a fuel stop, Qadvo replies with that item's most urgent next step — an overdue service first, otherwise the nearest upcoming due (miles/hours or days). Items with no reminders get an offer instead of a guessed schedule: tell Q the cycle ('remind me every 5,000 miles') and it creates the interval. **Choosing whether to keep that archive copy (v1.51.119):** the Transfer Ownership confirmation has a "Keep an archived copy of my records" checkbox, on by default. Uncheck it for a clean handoff and nothing remains in the sender's account after the recipient accepts — the tree, its logs, and its contents ledger all move to the new owner (ledger lines always travel with the item). **Scanning receipts for shared equipment (v1.51.81):** when someone scans a receipt for an item shared with them (editor/admin), the scan suggests merging into that shared record — an employee's parts receipt lands on the company's equipment instead of creating a duplicate; if they also own an identically-named item, their own copy is preferred. Viewer-role shares are never suggested (a viewer can't save a merge). Tap the **Shared** filter pill to see items shared with you or that you're sharing out. Share any asset (viewer/editor/admin access) from its detail page; financial data is hidden by default. The Overview flags shared items with a green users + arrow icon. When you invite someone by email, they get an invite email with a **one-tap sign-in link** — no password or sign-up form. The email says plainly that accepting activates a free Qadvo account for their address, and the confirm screen's Accept carries agreement to the Terms of Service & Privacy Policy (same consent as normal signup — the invite path doesn't skip it); until they accept, the account holds nothing and nothing is shared with them. They confirm who's sharing (and can decline if it wasn't meant for them, which notifies you), then land on the shared item. **Never share your password to give someone access** — invite them as an editor instead; everything they do is under their own name and access can be removed anytime. Invited people can set a password afterward, or sign back in later via **"Email me a sign-in link"** on the login screen. You do **not** need your own paid plan to view or help with items shared with you — only to create your own items. Adding or importing your OWN items needs your own plan; a shared viewer without one sees a "start a plan" prompt when they try (and can always leave a share from its detail page). **Ask Q is available to shared viewers** — if you're helping on a workspace someone else pays for, you can ask Q about the items shared with you at no charge. If your plan lapses, your own items go read-only but you keep helping with shared ones. **Transferring ownership (v1.51.156):** Share → Transfer sends the recipient an email with a **one-tap sign-in link that opens their Qadvo Inbox**, where the transfer waits with Accept Ownership; nothing moves until they accept. Someone not on Qadvo yet gets the same email — accepting activates a free account for that address (the old flow only emailed an account-creation link, and someone who already had an account got an Inbox notification and no email at all, so a transfer could sit unseen). The sender's Share screen also shows the transfer as pending until it's accepted or cancelled. **Cancelling a transfer or removing a share (v1.51.157)** clears the pending request from the recipient's Inbox and leaves a read-only notice in its place ("Transfer cancelled", "Share withdrawn", or "Share removed") — they never see a live Accept button for something that no longer exists. Sending from a development server puts a localhost link in the email; the Share screen warns about it, so send real transfers from qadvo.com. #### Archive **Archive is a hide switch, not a life event.** Archived items are hidden from the main view, the Due list and reminders, and their full history is preserved — use it for duplicates, mistakes, test items, and records you simply don't want to see. **It is NOT how you record that something was sold, scrapped, lost, or died — that is the item's Status (next section), which keeps the item's economics and comparisons intact.** **Archived children step aside (v1.51.130):** an archived item under a parent (an expired service agreement) leaves the parent's Contents list, stops counting in the parent's economics totals, and is no longer treated as coverage when you ask Q about the parent — it is still under the Archive filter, and Restore brings it back everywhere. Archiving is one-way down the tree: archiving a child never affects its parent. Use the **Archive** filter pill, or the "Archived only" toggles in the Overview tabs. **Where Archive lives now:** open the item → **Edit** → the **Archive This Item** row sits beside Delete at the bottom (two-tap: it turns amber with "Tap again"; tap once more within 3 seconds; archiving a group also archives its child items). An archived item shows an **Archived — Restore** pill in its header; tap Restore (two-tap) to bring it back — this works even on the read-only archived copy of a transferred item, which has no Edit button. #### Client work — the Own | Clients pill (v1.51.167, Business plan) **One account, two kinds of records.** A landscaper's account holds their own trucks and equipment AND their clients' properties, plants, and hardscape. **Every record has a side.** A record inside a parent takes the parent's side; a TOP-LEVEL record of any type — a group, a property, a single plant — carries its own mark (open it → Edit → the **Client work** switch, or turn the switch on when adding it; the line under it says how many records change sides on Save). Mark a top-level group and everything under it, at any depth, is a client's — properties, plants, photos, logs. Nothing on the records themselves changes. Inside a client group, Edit shows the switch on and locked, naming the group it inherits from. **The home screen then gets one pill that flips between Own and Clients** — it reads the side it is showing and takes the Archive pill's slot (Archive moved to Settings → Archive). **Own** shows the account's own items only: the grid, search, the Due list, the Overview, and Ask Q leave client work out, and a line under the pills says "Client work hidden" with a Show clients link. **Clients** (the pill turns orange) shows client work only, under a loud orange banner that says so; Ask Q reads client work only. The Client work switch appears on every record's Edit screen. On a NESTED record (one inside a group or property) the side belongs to the top of its tree, so flipping the switch asks which you mean: **mark the top record** (e.g. the whole property, everything under it) or **move this record out** to the top level and mark it on its own — with a Cancel; it is never simply locked (v1.52.14). The mode is remembered for the tab and lives in the address bar (?scope=clients). **An open item always wins:** asking Q about a client property with it open reads that property whatever the mode; a care guide or report on a named item is unchanged. **Move follows the side:** moving a record (or a whole subtree) under a parent gives it that parent's side — under a client group it becomes client work, under your own group it becomes your own — logs, photos, values, and history untouched, with a timeline note when the side changed. Moving a record OUT to the top level keeps the side it had (a plant taken out of a client property stays client work, with its badge), so nothing changes sides just because it was unfiled. The Move picker offers the client groups from either side. An item under a client group shows a **Client work** badge in its header; tapping it opens Move. **Marking needs the Business plan; reading and clearing never do** — an account whose plan lapsed still sees its client work and can move records back. **Maintenance account (v1.51.168):** client work is quiet by default — its timers never reach your Due list or reminder emails. A client property you maintain gets a **Maintenance account** switch in Edit (shown on properties on the client side); on, its care dates enter your Due list and reminder emails, and it carries its own visit cadence (spring & fall, spring, fall, or your own words) that the care guide prints instead of the business profile's. Off again and it goes quiet. **Reports follow the side (v1.51.169):** the Expense / Tax export and both backups in Settings cover the side the home screen is on — Own items or Client work — and say so (a scope line on the PDF, an Account Side column in the CSV, the side in the backup's file name and summary), so an insurance or tax report started from Own never lists a client's plants. Take a backup from each side for the whole account. The Lab Calibration report moved out of Settings onto the item's **Reports** button: it appears on any record whose contents include lab equipment (a lab property, a group of instruments, one meter) and covers that record and everything under it, whatever the mode. **A plan per property (v1.51.171):** a client property's maintenance plan is a **service agreement** — its own terms (plan name, price, billing: annual at enrollment, monthly, or per visit; cover line, what it includes, discount note, visit schedule, notes only you see) and a status: **offered** (a care guide went out carrying these terms), **active** (the customer said yes — the property is on your maintenance program), **paused**, or **ended** (kept as history; a new plan starts a fresh record). Open the plan from the property's **Edit → Maintenance account → "Set up / Edit the plan"** (opens on Active) or from the Care Guide builder's **"Price this job"** card when the offer is on (opens on Offered); the drawer shows the property's plant count, value under care (plants and hardscape only), last visit and next visit. **Plan tiers** in Settings → Business Profile: up to three bands by plant count, each with its own terms; the drawer prefills the band that fits the property and everything stays editable. Publishing a guide with the offer on records an Offered plan with the printed terms. The plan stays with you when a property is shared or transferred; the customer only ever sees the frozen guide. **The Clients tab (v1.51.172):** on the Clients side, the Overview's last tab is your maintenance-accounts board — one row per client property with its client group, plan and price (or "Start a plan"), status, last visit (newest maintenance or service log under it), next visit (from the plan's schedule, active plans only), plants, value under care, and the work due under it listed line by line (overdue red, coming due amber; each line opens its item). Tap the plan to edit it; totals run across the top. The Overview also carries a Switch to Clients / Switch to Own button. #### Credits (refunds, returns, rebates, reimbursements) Log a credit as its OWN entry — never by lowering the original charge, which stays the record of what was paid. In the Log form, switch **Charge → Credit** above the amounts and enter the amount as a positive number (and its tax, if the refund included tax); it saves as a negative amount. Pick the tab the money belongs to (a refund on a repair goes on Incident, a warranty reimbursement for maintenance on Maintenance) so it reduces that category. Optionally pick **Against** — the charge it offsets — and the timeline shows "Credit −$45.00 · against ". A credit reduces the item's operating cost, its $/year and 12-month trend, the Overview → Economics totals, the Excel backup's spend, the tax-preparation report (as a negative line), and it prints on the history and service reports as a credit. A scanned refund receipt whose total is negative opens as a Credit automatically. On the timeline a charge's amount is a neutral grey chip and a credit's is green, so money back never looks like money spent. The **Record** type (a claim, a deductible, a lender statement, an assessment notice, damage) shows as a document icon in grey — it is the paperwork of owning a thing, not an alarm — and is called Record everywhere, including the history report. #### Item status (lifecycle) — sold, traded in, scrapped, lost, died, retired (v1.51.149) Every physical item, plant, animal, and property has a **Status** button in its header action row (it replaced Archive there); the button's label IS the current status — "ACTIVE" normally. Tap it to open the picker. The choices depend on what the item is (a cow can't be scrapped, a tree can't be traded in, a QDoc has its own coverage status and no Status button at all): equipment, vehicles, and items offer **Retired, Sold, Traded in, Scrapped, Lost or stolen**; animals offer **Sold, Died, Rehomed / given away**; plants offer **Sold, Died, Removed**; property offers **Sold**. Choosing anything except Retired opens a short form: the **date** it happened, the **money** where it applies (Sale price / Trade-in credit / optional Salvage value), the **other party** (Buyer / Dealer / Scrap yard / New home), an optional note, and — for a trade-in — an optional **"toward" picker** naming the replacement item, whose Economics then shows the credit on its cost line. If the item has child items, the form lists them with checkboxes (all ticked) so you choose which go with it — a flatbed goes with the truck; anything unticked stays put. A contents ledger goes with its parent (promote or move a line first if you want to keep it); when a **property** sells the form says the opposite — contents usually stay with you — and offers to move them first. Once the container is sold, its ledger is **closed**: the lines stay visible as history with a lock banner, count in no total, and can't be scanned, edited, added to, removed, or promoted (the server refuses too). The one thing still allowed is **moving kept lines out** to a live item ("I kept the piano" — noted on both records); nothing can be moved *into* a sold container. To change anything else, bring the record back into service (Status → Active, with a reason). **The sale is written to the item's timeline as a Disposal entry** — that entry is the record; edit it from the Status button (the log pencil and trash refuse it), not from the log editor. **Retired** means still owned but out of service — no form, optional note. **What each status does.** Retired and every disposition leave the **Main** tab, the **Due** list, and email reminders (their timers stay in the data but never fire); they stay under **All**, and search still finds them. Sold / traded in / scrapped / lost / died / removed / rehomed also leave your **inventory value totals**, the **Insurance Inventory and Condition reports**, and any **public share link** of the item (which reads "no longer active"); Retired stays in value and in those reports because you still own it. A QR sticker on a sold machine no longer resolves. **Economics on a sold item shows a Closed ledger** — Sold for / You paid / Operating / Improvements / **Net cost of ownership** (purchase + operating + improvements − what you got for it) and $ per year over the years you actually owned it, ending at the sale date. **Ask Q** keeps disposed items in comparisons and cost questions ("compare my mowers' cost per hour" includes the one you sold, tagged) but leaves them out of "what do I own", "how many do I have", and what's due — they appear on a trailing "No longer in service" line; Q never infers a sale from a note. **Bringing an item back** (Status → Active) is allowed but asks for a reason; the disposal stays on the timeline marked as reversed and its proceeds stop counting. Cards on the home screen show a small pill over the item's photo for any status other than Active. The emoji **Badge** is different: it's a decorative tag you choose (a cow that calved, say); status is structured state with the consequences above. **Many at once (a herd sale, a lot of equipment).** A group has no status of its own, but its header carries the same **Status** button in the same slot; on a group it opens the **Contents** tab in picking mode (the tab's own **Set status** button does the same). Tick the items — at any depth, so a calf nested under its mother is ticked on its own — or use the quick-picks: "All", "Top level" (only records directly in the group), "Nested" (only the records under those — the calves), or by kind ("all Cattle (40)"); a quick-pick that ticks nested records opens their parents so nothing is ticked out of sight, and a collapsed parent shows a green count of what's ticked inside it. Then **Set status…**; **Cancel** leaves without changing anything. The picker offers only the statuses every ticked item can take (mixing calves and a truck narrows it to what both allow, and says so). For a sale you price it **each** (every item gets that price) or as a **lot total** (one check split evenly; each item's Disposal entry records its share and the lot it was part of, e.g. "share of a 3-item lot sold for $3,000"). Items already sold are skipped and reported, not failed. Only what is ticked changes — never a cascade: ticking a cow does not touch her calf, and ticking the calf does not touch the cow (the per-child bundle — "the flatbed goes with the truck" — lives on the single item's own Status form). The Contents tab also has **Expand all / Collapse** to open or close every nested level at once. A group itself has no status; it takes its meaning from what's in it. #### Inbox / notifications The **Inbox** (top-right) shows notifications — share invites, system alerts, reminders; the badge is your unread count. Tap one to act on it. **The reminder email (v1.52.47):** when timers come due, one email lists them. It may also carry **Something worth knowing** — one item with something new to say: a pattern across months, a run of repeated work, something never done, a comparison with things of the same kind, or what's known about its make and model from a web search, with the source named. Its headline is the email's subject line. Every figure in it — costs, how many times something was logged, readings — comes from the app's own calculations, never the writer's arithmetic. Before it is sent, every claim is checked against the item's record and the research (dates, spans of time, counts, causes, sources, general know-how stated as fact); a note that can't be backed after five tries is left out and the email goes without one. It never retells a reminder already listed in the same email. It doesn't bring up work logged in the last two weeks — the owner already knows it — unless that log holds something easy to miss, like a reading out of range; it never asks whether a fresh repair held, never nags about a step the owner said they'd take, and doesn't repeat itself: an item comes back only when something new has settled on its record (a log at least two weeks old). Each email also carries a **Did you know?** tip — one Qadvo feature and how to use it, taken from this manual — a different one each time until all have been sent. Since v1.52.48 the note's point is always something the item's own logs don't say (the same rule as Surprise me), and when an item has nothing new the email tries another before going without one. A timer is listed once in the email, and once in Q's "what's due" answer, however its name is capitalized. #### Settings & account The **Settings** page (gear icon, top-left) holds everything outside the daily flow: **Reports** (full Excel backup, per-asset history reports, lab calibration reports, crop-year P&L), profile / password / email alias, notification preferences, card warranty benefits, and plan & billing. Password fields everywhere (Settings → Security, sign-in, reset, invite) have an **eye icon** that shows or hides what you're typing. Qadvo **cannot show your current password** — it is stored only as a one-way hash, so nobody can read it back; signed in, set a new one under Settings → Security (no current password needed), or use **Forgot password** on the sign-in page. Sections are grouped top-to-bottom: Tools, Preferences, Smart Features, Account. **Default new items to Business** (Settings → Business) is saved to your ACCOUNT, so it follows you to every device (it used to be per browser, and a second device silently reverted to personal). Changing it takes two taps — tap, then “Tap again” within 3 seconds — like Delete, because it silently decides the tax treatment of everything added from then on. It seeds new items; **imported invoice logs inherit the item's own business flag** and only fall back to this setting when the item has none. The setting belongs to the Operation and Business plans: on a plan without it (or after a downgrade) it reads as OFF everywhere and new items stay personal, and the earlier choice returns by itself when the plan comes back. #### Expense / Tax report (Settings → Reports) The yearly **Expense Report (Business only)** PDF lists every business item at ANY depth under a group — a property → Trucks → Truck #10 → its service logs all print, each level indented, and a group's sub-total includes everything beneath it. (Before v1.51.151 it reached only one level below a top-level group, so a nested fleet printed almost nothing.) An item's flag is what counts; a log with no flag of its own inherits the item's, and a log explicitly marked not-business is left out even on a business item. The CSV export is the same data as a flat list. **Loan interest is on it (v1.51.155):** a lender statement logged as a Record with no cost contributes a "Loan interest" row for the interest amount (the payment's principal is not an expense and never appears); the PDF's totals say how much of the total is interest, and the CSV categorizes those rows "Loan Interest". The rule is the same one the scanner logs by: a loan payment is principal (already in the item's purchase price) plus interest, and only the interest is an expense — so a log that carries loan figures is counted by its interest even if a cost was typed on it. Every item's header states its tax treatment as a chip next to its type: **Business** (blue) or **Personal** (grey) — never silent, so a mis-flagged item is visible before it goes missing from the report. Change it in Edit Details. Coverage documents, cards, and receipts carry no tax flag and show neither. #### Care guide — the service pass (v1.51.176) Right after the cover, a **This month** page says what the yard needs now, worked out from the records: establishment watering for anything planted in the last three months (deep soak about twice a week) or its first season (weekly in dry spells), the calendar's tasks for the current month (trunk guards before rut, pruning), blooms to enjoy, and next month's work coming up. Beside it, **"Something wrong with a plant?"** is the guide's most prominent action — on screen it sends a note with a photo to the landscaper; on paper it prints as the sentence with the business email. The guarantee section opens with **If something goes wrong** in three lines (tell us, send a photo and the plant's name, we reply with the next step — and the watering guidance the terms refer to is this guide), then **Your guarantee, in short** — two customer lines (what's covered, what voids it) that Q drafts from the operator's pasted terms via **Settings → Business → What your customer needs to know → Draft from my terms**, which the operator edits and saves — then the **full guarantee terms**, the operator's own words, never rewritten. The cover's value box states that the guarantee covers the plantings figure and hardscapes are not included. "We can handle this" prints only on calendar months that hold a task, never on a bloom or harvest. The plan price prints with its period ($79/year, $79/month, $79/visit) from the agreement's billing term. A review incentive note may say `{plan}`, which prints the plan on that guide. Visit focus and the site-plan caption say whether trunk guards go on (before rut) or come off (spring). The guide covers only what is still in the ground: a plant or hardscape marked sold, died, removed, retired or archived is left out of the cards, the calendar and the guarantee, and the builder's checklist never offers it (v1.51.194). The claim path's first line carries the **Something's wrong with a plant** button itself and a click-to-copy business email; its second line links to the planted table. The living-record block (the private QR for a shared customer) adds: something wrong with a plant? add a photo to its record right there — the same record the landscaper works from — then tell us; that line prints only when the customer's share is editor or admin (a viewer can look, not add). Nothing alerts the operator when a customer adds a photo, so the guide always says to tell them too. In **Create care guide**, if the property's only person-share was declined or expired, the share-first note names who and when, so the operator knows to send a fresh invite. **Tiers and plans (v1.51.178):** a tier in Settings is a template for NEW plans; editing it never rewrites a plan a property already has (an active plan is a price the customer agreed to). When a property's plan differs from its tier, the plan drawer (Edit plan) shows a line naming both — the tier's current terms and what the property has been on since when — with a **Use the Settings values** button; nothing changes until saved. Billing defaults to annual; per visit must be chosen. A review-incentive note that names a plan is rewritten once to `{plan}` on the next profile load, so it prints the plan of whichever guide it is on. In **Create care guide**, editing and saving the plan keeps every other choice in the builder (photos, toggles, recipients) — the builder resets only when it opens (v1.51.179). #### Starting a customer's plan, and what the customer sees (v1.52.26) When a customer taps **Request this** on the care plan in their care guide, it arrives in your **Inbox** as "Care request: Care plan" (with their name when the guide was issued to them) and as a locked note on the property's timeline. The request doesn't start anything — you decide. The request carries a **Start the plan** button: it opens that property's plan already set to **Active**, with the terms that went out on their guide; one save starts it (the timeline notes "Maintenance account on", and the property's care dates reach your Due list and reminder emails). The request then shows "Plan started". You can also start or change a plan from the property's **Edit → Maintenance account**, the care guide builder, the **Change** button on its Maintained by line (your own account only — a customer never sees it), or right on **Overview → Clients**: each property's **Status** is a picker (Active, Paused, Offered); ending a plan is done from the plan's name, which says what ending does. A paused plan shows in amber everywhere — on the Clients tab and on the records' Maintained by line ("Paused · on hold since …"). Once a plan is active (or paused), every record under that property shows it on its **Info** tab, for you and for the customer on their shared records: **Maintained by** — your business name with the plan's state (Active, or Paused in amber), the plan's name, its visits, since when, and the next visit (a paused plan says since when it's been on hold; an offer that wasn't accepted shows nothing). Each **plant** still in the ground also shows its **Plant guarantee** — through the date your care guide gives it (one year from planting, from its planting date or else the day it was added), or ended — but only when you've written guarantee terms in your Business Profile; Qadvo never states a promise you didn't make. Prices never show on these lines. A maintenance plan isn't insurance: it doesn't count in the coverage check. **Transferring** a property ends your maintenance plan on it (the transfer screen says so first — to keep maintaining it, share it with the customer instead); a customer claiming their record from a care guide is a share, never a transfer, so it never ends a plan; and records Qadvo's onboarding team delivers to you carry their plans with them. #### Business Profile (Settings → Business) In **Settings → Business → Business Profile**, set your logo, business name, address, phone, email, website, and a default guarantee — once. (Before v1.51.190, revoking the last published copy that carried your current logo could delete the logo file itself, so later guides showed a broken image; a logo is now never removed by revoking a report. If a guide's cover shows a broken logo, upload the logo again in Business Profile and generate the guide again.) Every customer-facing report (and future invoices) uses this branding automatically, so you never re-enter it per report. On the care guide the phone, email, and website print in the header and the closing line; on screen the email is click-to-copy (not a mail link — those misfire for webmail users), and it prints as plain text (v1.51.147). This is part of the paid Business plan. The same screen holds the care-plan defaults: scheduled visits (spring & fall / spring / fall — or your own wording), the **Plan tiers** (bands by plant count that prefill a property's plan), and an optional review link with an incentive. The single "Care plan offer" template (plan name, price, cover line, what it includes) was retired in v1.51.178 — every property's plan is its own (Edit plan on the property, or Price this job in the care guide builder), and a guide prints that plan or no offer at all. Every word is your own — Qadvo prints it verbatim and never invents an offer or a promise. #### Hardscapes (patios, walls, edging) Hardscape — built landscape construction — is a **modifier on the Item type** (Equip is machinery): Paver Patio, Retaining Wall, Walkway, Steps, Bed Edging, Driveway, Fire Pit. **Scan a photo** and Qadvo identifies the material CLASS (concrete pavers, natural flagstone, limestone block...), laying pattern, a size estimate, and visible condition — it will NEVER guess a brand or product line (not readable from photos) and never auto-creates a warranty for a material class. Cards carry an upkeep rhythm, PROTECT IT do-instructions, and normal-for-this-material notes. The Info tab is tap-to-edit (surface type, material, installer, installed date). **Timers**: the wizard offers Paved Surfaces (wash/joint sand/re-seal), Retaining Walls (inspection + drainage), and Bed Edging (frost-heave re-set) collections. **Care guide**: when a property's tree contains hardscape, the guide becomes "Your Property Care Guide" with a Hardscapes & Surfaces section (value-sorted) and spring wash/inspection calendar entries. A guide with hardscape and no plantings skips every plant-only part: no plantings count, no "What's in your yard", no Everyday care or Watering 101, no planted record table; it keeps the cover, the hardscape section, the upkeep calendar, and the operator's guarantee terms if written. **Value**: the AI estimate prices hardscape as an INSTALLED construction (materials + labor from material class × scanned size); a hardscape can't be sold separately — its resale value is realized as property value. **Economics**: standard asset view (cost of ownership from logged maintenance; depreciation appears when the item is marked a Business Expense — app-wide rule). #### Service Report (BETA) On any item with log entries: **Reports → Service Report** — the contractor leave-behind. The builder opens with the **latest visit pre-selected** (every entry sharing the newest entry's date); pick or unpick entries, optionally add a headline ("Fall furnace tune-up"). **Costs are off by default** (toggle on; the invoice is a separate document — internal system entries are never included). The report is a compact 1-2 page branded record: equipment identity + photo, each entry with description and the photos captured during the work, verification QR — handed over as a link or scanned at the door. **"Clean up notes for the report"** (on by default): AI rewrites field shorthand into customer-ready language under a strict rule — NEVER ADD (no invented inspections/findings/work) and NEVER DROP (every action, part, outcome, and measurement survives). The owner previews and can edit every rewrite before publishing; the original timeline logs are NEVER modified — the record stays what was actually typed, the report is what was chosen to present. If the equipment is shared with the customer, their private living-record QR is included (the builder nudges to share FIRST — snapshots freeze at publish). **Every published report** (all types) also writes a LOCKED entry into the item's timeline with a "View this report" link to the exact frozen copy — reports stay retrievable from inside the record itself. **Revoking** a published report (from the item's Reports list) writes a matching locked "revoked" entry naming the report kind and the date it was published (v1.51.147), so the record always shows that a report existed and when it was withdrawn — a printed QR that no longer resolves is a withdrawn report, not a mystery. Revoking is permanent: the frozen copy is deleted and its link stops working; to correct a report, regenerate it instead (the old link then points to the newer edition). Revoking now asks **why** (a short reason, required) — it is written into the revoked entry in your own words, the entry names the person who revoked it, and the earlier "published" entry's link greys out with a strike-through so nobody clicks into a dead page. Published and revoked entries both name the person (inside a team workspace, the member who did it, not the account) and carry the copy's number (#1234, the same on both), so a published/revoked pair can be matched by eye in a long log; the published entry itself shows the revoke date and reason. New published or revoked entries appear on the open item's Activity tab at once, no reload. #### Rental Condition Report & Insurance Inventory (BETA) Two reports on the **Reports** button of properties, groups, and items with contents (both labeled Beta while they mature). **Empty properties qualify too (v1.51.78):** a property with no items linked yet can generate both reports — they document the property itself (its photos with dates/captions, overall condition, and notes, including a room scan's "Contents observed" listing). That's the move-in walkthrough of a vacant rental: photograph the rooms, scan, set the condition, generate. A "Property — overall documentation" section also opens the report on furnished properties; the property's own value never counts toward the contents total. **Rental Condition Report**: room-by-room move-in/move-out documentation — each item prints with its photos, a short DESCRIPTION (what it is + basic care, drawn from the item's Notes, boilerplate stripped), and its CURRENT CONDITION; dollar values hidden on tenant copies unless toggled on in the builder. The Notes field doubles as the place to document move-in observations — free text prints as the description, and a line typed as "Condition: " surfaces as an observed condition. (The same item description now also shows on the Insurance Inventory report.) **Insurance Inventory**: room-by-room contents with replacement values (each line states its basis — replacement, current, or paid), room subtotals, grand total, attached policies with coverage limits, and active warranties; values always on. **Rooms are groups**: put items in a "Kitchen" group under the property and the report organizes itself; direct children land in "General". Plants count as contents (a tenant can kill a tree); people/animals don't. The builder has room-level checkboxes to include/exclude a whole room. **Current Condition** is set on any item (Info tab tap-chips or Edit Details): five grades (excellent/good/fair/poor/damaged) + an observation note; the as-of date is SERVER-stamped when set and can never be back-dated; unset items honestly print "Not assessed" — a grade is never invented. **Photo provenance**: every new photo carries a server-stamped upload date, and captions added in the photo viewer print under the photo on reports. **Frozen evidence**: deleting a photo referenced by a published report keeps the file so the frozen report never loses its evidence (the app says so when it happens); revoking a report cleans up files nothing else references. Both reports publish as frozen snapshots (verification QR, supersession, copy-link, revoke). These are documentation the other party can weigh — never presented as a legal instrument, appraisal, or statement of insurability. #### Care Guide reports On **any item or group whose tree contains plants or hardscape** — a single tree, a group of plantings, or a property with plants/hardscape nested at any depth — tap its **Reports** button and choose **Create Care Guide** to build a customer-ready report. The guide includes a **Watering 101** section (soil-check method, slow-deep watering technique, a volume reference table, watering-bag guidance); each plant card states **its own water amount** from its type and size, and recently planted items add the **establishment watering schedule** from their planted date. **Lawns**: the "Lawn" plant type covers seeded or sodded turf areas — its card gives install-method-specific establishment watering (sod vs seed, detected from the notes), and lawns get a Lawn & Turf timer collection (fertilize, overseed, aerate, mower blade sharpening). The care calendar includes **protective care** for young trees (under ~4" caliper): fall trunk guards before deer rut, a pre-snow rodent-girdling check, and spring guard removal. The cover can show your **region and hardiness zone** — set Service Area and Hardiness Zone in Business Profile (the AI does not know your location; plant timing knowledge assumes a Northern-Hemisphere temperate default). The cover counts **unique plantings**; the record table carries total specimen counts and values including quantity multiples. In the Reports screen, each published copy has a **copy-link** button. **Winter-hardiness verdicts**: once a Hardiness Zone is set in Business Profile, plant scans add a zone-aware verdict to each card, and the chip says in words what its color means; a one-line key at the top of "What's in your yard" explains all four (v1.51.190): **Hardy here** (green), **Protect in winter** (amber — borderline in your zone; shelter and mulch), **Bring indoors for winter** (rose — will not survive outdoors here; a tropical sold as an "annual" belongs in this group), or **Cultivar not set** (gray — hardiness varies by cultivar and the record doesn't name one; that is a gap in the record, not a warning about the plant). The care guide builder lists the plants with no cultivar set before you publish, so you can add the cultivar and re-scan. A plant's botanical name is printed as a name only; when a scan was unsure ("likely a Northern Lights hybrid"), that hedge prints as a separate *Identification note* under the name, never inside it. Scans made BEFORE the zone was set don't have verdicts — re-scan the plant to add one. (The button only appears on groups with plants.) In the builder you pick which items to include, whether to include the guarantee, and — optionally — a property photo and an overhead/site-plan photo from that group's gallery. The report has a cover page, a "what's in your yard" section with each plant's identity and care facts — including a wildlife verdict in plain words (e.g. "Deer and rabbit resistant"), a PROTECT IT note giving the species' top preventable killers as positive do-instructions, and a NORMAL FOR THIS PLANT note for traits that look alarming but aren't (these appear after a plant has been scanned or re-scanned; each card uses the plant's chosen cover photo) — a month-by-month care calendar built from those specific plants, and a plant-value table (each plant's estimated value with a running total, shown as an insurance reference — explicitly not a statement of warranty coverage). **Publishing & sharing:** when you generate the guide, Qadvo saves it as a snapshot and gives you a shareable link; the cover shows a QR code the customer can scan to open the guide on their phone. When you text or email that link, the preview card the phone or inbox draws shows the guide itself — *Your Property Care Guide — [property]* with the cover photo you chose — not a generic Qadvo card (v1.51.189); the photo is fetched fresh each time a preview is drawn, so an old text's preview never goes blank. Published guides are hidden from search engines (the link is the only way in). A published guide is a frozen copy labeled with the date it was prepared — it does not change after publishing. You can also print it or save it as a PDF. The guarantee section shows only YOUR own terms from your Business Profile (with formatting like bold, links, and lists preserved); if you haven't set guarantee terms, that section is omitted rather than filled with a generic placeholder. Branding comes from your Business Profile. **Give the customer a live account:** if you've shared the property with your customer's email BEFORE generating, the published guide also includes a second QR at the end — their private "living record." They scan it, get a one-tap sign-in link to the email you shared to, and land on their own always-current view of their plants, where they can add photos and ask Q about any plant, free. If you haven't shared the property yet, the builder points you to the Share button first (the snapshot is frozen at publish, so share, then generate). (Part of the Business plan.) **Selling mode — the care-plan offer (v1.51.146).** When the property has a plan, or the profile has plan tiers to price one from, the builder shows an **"Include care-plan offer"** box — OFF by default on every guide, so an insurance or certification copy never carries sales copy; tick it only for guides meant to sell. A selling-mode guide adds: a **value est.** box on the cover — plantings and hardscapes as separate lines, each the estimated replacement cost of what's valued (whichever is present shows; when both are, a landscape total with the split beneath it; not a statement of what the warranty covers) with your cover line linking to the offer; an **offer block** right after your warranty terms (plan name, price, what's included, discount note, and a computed "next scheduled visit" — e.g. "September 2026 — trunk protection on 15 trees" — worked out from that yard's calendar, always for a month still ahead of the reader); a sticky offer bar on screen that prints as a footer line on every page; a "We can handle this" request line on each calendar month; **request buttons** on every plant card, month, and the offer; a computed site-plan caption; and a **review ask** at the end when a review link is set. The offer is frozen into the published guide like everything else — changing your price later never alters a guide already in a customer's hands. **Request buttons go to your Qadvo inbox, not email:** when a customer taps one (optionally adding a note), the request is written to the property's timeline as a locked **Service request** entry — the permanent record, which survives clearing the inbox — and an inbox notification alerts you (the badge updates within about 30 seconds). When the guide was published for a specific customer (the living-record share), the request names that person; otherwise it says "a customer". Guides published before the offer existed, and guides generated with the box unticked, show none of this. On screen, plant and hardscape cards are collapsed to one summary row each — tap a card to expand it, or use **Expand all / Collapse all** in the section header; print always shows full cards. The cover QR (scan to open the guide on a phone) is the same on screen and paper — a guide open on a tablet at a show table can be scanned by walk-ups. #### Qadvo Service Card (the dry-erase magnet) The Qadvo Service Card is a printed dry-erase magnet for an equipment cab or shop wall: an EQUIPMENT line, a handwriting table (DATE | HRS/MI | SERVICE | BY), NEXT DUE and NOTES at the bottom, and a QR code that opens Qadvo. **Use it like any maintenance card, then "snap it before you wipe it":** photograph the card with **Ask Q → Photo**, the **+Log** scanner, the item's gallery **Scan with AI**, **Add & Scan**, or **Add Item → photo** — every scan entry recognizes it. Qadvo reads the handwriting, matches the EQUIPMENT name to your item (or an item shared to you with editor access — a crew member's snap logs to the owner's machine), and opens a review screen. **You confirm before anything is saved** — every cell is editable (fix a misread initial), and because the SAME card gets re-snapped each time a line is added, Qadvo flags each row: new entries are included, rows already in the record are skipped, and close calls are marked "possible repeat" for YOU to decide — Qadvo never silently drops a row it isn't sure about. Each confirmed row becomes a maintenance log entry with the reading recorded on the item's tracking metric (odometer/hours), the BY initials preserved, and the card photo attached as evidence. **NEXT DUE becomes a real countdown**: one tap in the review creates a service countdown at that reading (named from your NOTES — e.g. "Brakes"), which alerts like any Qadvo interval — and if the target is already behind the newest reading, the review says so with both numbers instead of creating an expired timer (raise the target or service it first). **A row that names an existing countdown resets it**: writing "Oil/Filter/Brakes" with a reading restarts the Brakes countdown at that reading, same as the +Log reset rule. The card photo attaches to the newest entry as evidence (one checkbox; skipped with a note when the photo is already in the item's gallery). Wiping the card when full needs nothing in the app — the record is already saved, and re-scanning an unwiped card just re-skips the old rows. #### Table reader — a shop board or a card statement into many logs (v1.51.173) A photo of a **shop whiteboard** (one row per piece of equipment: name | current miles or hours | next service due at, or a note like "serviced @740 hrs") or a **bank/card activity screenshot** (date | payee | amount) is read as a table of rows, through any of the existing doors — Ask Q's photo, Add Content, or +Log on an item — and opens its own review; nothing is written until Save. **Board:** each row becomes a reading log on the record it names. Shorthand is matched three ways — a pick you saved before (remembered per account), the name rule, then one AI pass — and every row shows its pick in a searchable picker (records another row already holds are listed but disabled). The board's header date dates every entry. Its due-at column builds a countdown per row; the **interval length is per item** (typed on the row, else the item's existing timer, else the unit's default — never overriding an item that has its own), so the timer shows what is already used, even past due; a "serviced @ X" note works too (last service + length). A re-snapped board skips readings already logged and leaves a timer alone when it already says the same thing; a changed due-at refreshes it, never duplicates it. **Statement:** scanned on a business or group, each charge is filed by payee onto the record under it — a vendor's own record ("Anthropic services") or a kind-of-spend record ("Advertising") several payees fall under; a payee no record claims stays on the group with a "pick a record" chip, and a pick is remembered. With no record named, the remembered payees decide the destination (the top business group they share). Rows already logged anywhere in the account are flagged and skipped by default. Each entry stores the total paid as its cost and carries its own copy of the screenshot. New records are never created by the reader — add them from Add Content. Groups carry the **Business Expense** toggle like any record, with an option to mark everything under them at once. #### Hand it to Q — any paper or photo, filed where it belongs (v1.52.29) In Ask Q, tap **+ → Photo** or **Upload a photo or PDF** and hand Q any paper: a shop invoice, a store receipt, a policy, a title, a manual, a purchase order. Q reads it and shows a **filing card** before anything is saved: what the paper is, what it will create (a **Log** with its tab, a **QDoc** for a policy or document, the item's **Info**, a **Photo**), the papers themselves (tap to open), the record it goes on and the facts it read. Nothing is written until you tap **File it**; **Dismiss** discards the upload. - **Where it goes:** Q finds the record by something unique on the paper (a VIN, serial, plate, unit or policy number), by the item your note names, or by the item's page you're on. A name read off the paper counts too, and the card then says so ("Found by the name on the paper: check it's the right record"). **A paper that names only the kind of thing (v1.52.37)** — a trailer repair that doesn't say which trailer — goes on the group you keep those in (a "Trailers" group), or on the group all the matching records share when it is mostly them, and the card says so ("It says trailer but not which one, so it goes on Trailers for now (12 there): pick the one if you know it, or move the log later"); a payee's remembered record still comes first, and when the matches sit in different places Q asks instead. When Q isn't sure what kind of paper it is, the card says that first. Q never files a purchase price as a cost, updates a policy already on file instead of adding a second one, names a renewal after its reminder so the reminder resets, and treats a paper whose VIN or serial isn't the record's as another item. - **Store receipts:** only the item's own lines count, with its share of the receipt's printed tax; the card lists the lines it counted. - **Credits (v1.52.37):** a paper that gives money back — a refund, a return, a rebate, a reimbursement, a credit memo, or a correction / credit ticket against an earlier invoice — is filed as a **credit**: the card's **This paper is: A charge | A credit (money back)** switch is set to credit, the amounts stay positive on the card and are saved negative (as +Log's Credit switch), so it lowers the record's cost instead of adding to it. **It reverses** offers the record's charges, the one carrying the invoice number the credit names first, to link it to. The card always asks about a credit; when the credit names its invoice number and exactly one charge on the record carries it, that charge is linked already. The bulk invoice importer does the same (a **Charge or credit** box on each row). Every other way money comes in reads credits too (v1.52.37): **+Log → Scan** of a refund or credit ticket (a PDF or a photo) opens with **Credit** already selected; a **forwarded credit email** waits in the Inbox — it never files itself, even for a payee set to auto-file, and never becomes a new item — with the same **A charge | A credit (money back)** switch on its card, and filing it links the charge its invoice number names on the record you pick; a refund row on a **bank or card statement photo** (a minus sign, parentheses, or CR) reads as a negative amount and files as a credit (its box is labeled "Credit (money back)"); a **voice log** of money back ("got $40 back for the returned part") saves as a credit and is flagged for review. A minus printed after the figure ("1,204.50-", the way dealer correction tickets print it) counts as a minus. In the Edit Log screen, **Re-scan file · AI** (it re-reads the log's own files and saves what it fills) fills a paper that gives money back as a credit, never overwrites an amount you saved (a credit included), says so when a credit paper sits on a log saved as a charge, and moves a billed paper saved as **Measure** to its kind of cost (Maintenance, Record…), since a reading has no cost; with several files, tap the files to read together (the whole row picks it), and a file attached twice is listed and read once. Before v1.52.37 a credit was filed as a cost; an old one can be fixed by opening the log, switching it to Credit and saving. - **Total:** the card shows the log's total (before-tax amount plus tax) as the timeline will, and the paper's own total beside it when they differ. When the paper is already on file, the card unticks it and its button reads **Already filed on **. A paper counts as already filed when its invoice number is on a log with the same date (the log's service or receipt date), or its date, amount and payee match one; a number repeated on a different date is an account number — a registration's plate number on every year's renewal, a lender's on every statement — so next year's renewal is a new charge (v1.52.38). - **Readings:** a quantity your note or the paper states ("log 100 lbs of salt") lands on the item's own metric (Salt, in lbs), so its reminders reset like any +Log. If the item tracks a metric the paper is about but no amount is given, the card shows an empty box to fill. A number in another kind of unit (a count of bags) is never forced into it. - **Two dates:** the **Receipt date** is the paper's own date, when it was bought. The **Service date** is when it was done: the date your note gives ("I added salt on the 25th", "9/25/2026", "yesterday"); else the paper's date when the paper records the work itself (a shop's invoice, a bill, a renewal, an inspection); else today, for things bought to use or install later (supplies, parts, materials). +Log's Scan follows the same rule, and the timeline shows a differing receipt date as "Receipt: ". - **Paid by** is filled from the receipt (the green Paid chip on the timeline), and the log takes the item's business or personal setting. - **Change** on the card: "It's actually…" lists the other kinds of paper with the current one ticked; **Just keep the file** puts the file in the gallery only; **Make a new item** opens the new-item review. - **After filing** you land on the record, and a notice says what was filed, with **Undo**: it takes the filing back and reopens the card with your papers. - **Add Item and Add Content** hand papers (PDFs, receipts, policies, purchase papers) to Q the same way; a photo of a thing still opens the new-item review. - **A purchase paper for several items** (one order for two trucks) makes one item each, with its own price, VIN or serial, mileage at purchase and its own copy of the paper. After **Create**, the app opens what was made. - **Every door files the same way (v1.52.30):** a forwarded email that isn't a receipt (a policy, a notice, a manual, a title) waits in the Inbox under **Papers to file** with its card, instead of becoming an item — that card is its one Inbox entry (no separate notice beside it), and the bell counts it until it's filed or dismissed; closing the card with its X keeps it waiting there — only Dismiss or filing it takes it out (v1.52.31) (a receipt keeps its Unfiled card). **The email itself is the paper (v1.52.35):** keeping a forwarded notice files it as a **Record** entry on the chosen record's timeline, dated from the email, carrying the email (open it from the entry) and any attachments; a log, document or policy filed from an email carries the email too. Before v1.52.35 an email with no attachment was "kept" as nothing — if one seems missing, forward it again and file it. **Add & Scan** or **Scan with AI** on an item turns a receipt into a log card and a policy into a policy card, instead of merging it into the item's details; a card dismissed after Add & Scan discards the upload. **+Log's Scan** hands a policy, a purchase document, or a paper whose VIN, serial or plate names ANOTHER of your items to the card, which proposes the right item — photos and PDFs alike; anything else fills the +Log form as before. A fuel receipt or a loan statement opens its own review (fuel's gallons and odometer; a statement's rows). A paper saying an item was sold, traded in or scrapped opens the item, where its **Status** button records it. An invoice partly paid by insurance or a warranty logs only YOUR share, with the invoice total and who covered the rest in the description. Q's own "Yes, log it" carries an amount you state ("log 100 lbs" of something the item tracks) onto that metric. #### Dates and times (v1.52.29) Every date and time Qadvo writes or shows — a log's date and time, notes, Q's "today", whether a policy is in force today, report dates — is in your device's own time zone. A log saved in the evening stays on today's date. #### Forwarded receipts → logs (the Inbox's Unfiled cards, v1.51.174) Forward a receipt email to your **save address**. It's shown, with a copy button, in **+ Add Item** at the top level (under **Or create from email**) and in **Settings → Inbound Email**; in a workspace only an admin sees the workspace's address. The whole email is read — subject, body, and attachments: a PDF's text, a scanned PDF's pages, or a photo (v1.51.180) — so a scanned invoice forwarded with a bare signature for a body is read from the scan, and a subject like "Invoice for " names the record it belongs to. **A word at the front of the subject says what to DO with the email (v1.52.10):** "log" or "add" means a log on an existing record — "log to Lake House" names the record, plain "log" leaves it to the picker on the Inbox card — even when the email is not a receipt (an approval letter, a notice, a confirmation); "new" means a new record whatever the email looks like. Plain words, no symbols. **A follow-up in the same email thread** (a reply or forward of a subject that already created a record, or that was filed from the Inbox, within the last four months) lands on that record as a log; the card says so and lets you pick another. Without a word or a thread, a receipt for a new thing still creates a record and a non-receipt is left for you to file. **When the receipt doesn't say what it's for, the forwarder says:** put the item's name in the subject ("Truck #7 White F150") or a line above the forwarded message ("log to Truck #7") — that instruction outranks anything on the receipt, and the charge lands on that record's card (with the picker open if the name fits more than one record). A repair, service, or fuel invoice that names nothing still lands in the Inbox with an empty picker — never as a new item; the card says to name the record next time. Only a purchase of a new thing creates an item — and that item's notes carry the order as the email states it: order number, seller, date, what was bought, the lines, tax and total (v1.51.182). **Already logged** is checked account-wide by date, amount and payee: an entry's amount counts as both its pre-tax cost and cost plus tax, and the receipt's as both its total and total less tax, so an invoice already imported pre-tax is recognised (v1.51.183); the card then offers to attach the receipt to that entry instead of adding a second charge. **Auto-file (v1.51.185):** with the payee's switch on, a receipt files itself when its amount is within half to double of ANY earlier charge from that payee on the record (a plan renewal and usage bills are each their own precedent); a card that still waits says why under the switch — the first charge from that payee on the record, or an amount unlike any on it. The remembered attachment rule keys on the file's kind (invoice.pdf vs receipt.pdf), so a dropped invoice stays dropped next month. An item created from a forwarded email takes the account's Settings default for business. **What it becomes (v1.51.186):** a durable good keeps its category (equipment, vehicle, property); a ticket, a course or book, a download, a service, a subscription or supplies is a plain **Item**. The receipt rides on the item's **purchase entry** (date, "Purchased — name", cost with tax beside it, vendor, invoice number, the PDF or the email body as its evidence) — the locked "Asset Record Created" entry is housekeeping only — and the header's **Receipt on file** chip opens the same receipt. If it reads as a **charge for a record you already have** — the payee was remembered from an earlier filing, the email names one of your records, or the receipt is a recurring bill (a subscription renewal, a usage bill) — it lands as a **card at the top of the Inbox**, not as a new item: date, payee, amount charged, tax if shown, description and log type read for you and editable; the record it belongs on suggested (a searchable picker offers the likely records, then all); the attachments listed with a **keep box** each (keep the receipt, drop the invoice); View receipt shows the email itself. Three ways out: **File it** (writes the log on the record, cost pre-tax with the tax beside it when a tax line exists, the kept PDFs and the email body attached as that entry's own evidence — an EMAIL chip on the timeline opens it; the body stays beside the files, and editing the entry later keeps the body, the invoice number and the files (v1.51.194) — and the payee → record pairing remembered), **Make a new item** (the email was a purchase of a thing: the create-an-item path, warranty and return research included), or **Dismiss**. Anything else forwarded still creates an item as before. **Auto-file:** on a card, tick "File receipts automatically"; later receipts from that payee file themselves — with the same attachments you kept last time — and post a **Filed** notice with **Undo** (deletes that entry, returns the card) and a **recurring** pill that opens the payee's settings: the record it files to, the switch, the dropped-file rule ("keep all instead"), **Forget this payee**. The same pill sits on every card. An amount far outside that payee's usual range waits in the Inbox instead; the first receipt from any payee always waits (auto-file needs one filed charge as precedent). A hand filing always wins over the remembered setup. **Repeats:** a receipt whose charge is already logged anywhere — from a statement scan, or the same email forwarded twice — comes as a card headed "Already logged on — " with **Dismiss** as the default and **Attach receipt to that entry** beside it (puts the receipt on the existing entry, adds nothing to any total); "File as a second charge" only when it truly is one. **The bell counts waiting cards**; a card is the notification — there is no separate note for it, and a Filed notice disappears on Undo. Attachments arriving with the email (PDF, images) reach the entry; the email's own body is kept as the receipt when there is no attachment. **What an emailed paper says (v1.52.57):** an emailed invoice's meter reading goes on the record's own meter, in the unit the record already tracks (the Inbox card shows it before filing and it can be changed there, or cleared to file none); and what the attached paper says that the log has no field for — the VIN or plate, the vehicle or machine as written, the unit number, handwritten notes and stamps, the shop's own guarantee, every line as printed — is kept with the log under **On the paper**, the same as +Log, the importer and Hand it to Q. A repeat attached to the entry that already holds the charge brings its paper's reading too, and fills that entry's meter reading only when it has none. #### Live records (v1.51.174) A record you have open updates the moment something lands on it — a receipt filing itself, a crew member's log on a shared item, a scan committed in another tab — with no reload and no polling: the server announces each change and open screens re-fetch. The Inbox and the bell update the same way, as do the grid, the Due list, and the Overview when a server-side process (a forwarded receipt, a table scan) changes your records. Your own actions refresh as they always did. #### Sessions sheets (the table reader's third mode, v1.51.174) A photo or screenshot of a **dated list of values for one thing** — a practice log ("Tango Practice: Date | Minutes"), a workout or feeding log, readings taken over days — is read as **Sessions**: scan it on the record (+Log → Scan, or Ask Q / Add Content with a note naming the record) and the review shows one row per line with its date and value, the **measure and unit set once** at the top (matched to a metric the record already tracks — "Tango" in "min" — else the sheet's title with the unit as written), and Save writes one **performance log per row**, the value as a tracked measure, exactly as +Log writes it, each with its own copy of the image. Two sessions on one day stay two logs. Repeats are **counted, not matched**: scanning the same sheet again skips every row already logged; a third same-day row is the only new one. Alongside the board (equipment | current reading | due-at) and statement (date | payee | amount) modes from v1.51.173. #### Whether a warranty is in force Qadvo works out whether a manufacturer warranty or a return window is in force from its **end date and today's date** — never from a stored label. If a record has the warranty's terms but **no end date** (common on older or second-hand items, where the purchase date or the term isn't known), Qadvo shows the terms and says plainly that whether it is still in force is **not known** — it will not call it active. To get a real countdown, add the warranty as a Countdown timer with its start date, or scan the warranty card. A warranty whose end date has passed reads as expired everywhere: Ask Q, reminder emails, the Coverage card, the history report and the PDF and Excel exports. **How the manufacturer warranty gets on a record (v1.52.15):** Qadvo researches it only for a manufactured product with a **known brand** — the brand you typed, the brand read off the receipt or the item, or the one Q identifies in Auto-fill. It never treats an item's name as a brand, and it never researches one for a group, a person, an animal or a plant (before v1.52.15 a cow or a folder could pick up an invented warranty this way; those were removed). A return window or a card benefit already on the record does not stop the manufacturer warranty from getting its own countdown. Forwarded receipt emails now record the brand too. **Vehicles (v1.52.52):** a vehicle's maker's warranty is "so many years or so many miles, whichever comes first, counted from the vehicle's first sale". For a vehicle you bought NEW, Qadvo counts the years down from your purchase date, and the terms note that reaching the mileage limit first ends the cover sooner — the countdown follows the years, not the odometer. For a vehicle bought USED — the "Original Owner" switch on the item is off, or it was bought two or more years after the model year in its VIN — Qadvo shows the terms it could verify and sets NO countdown: the years began at a first sale the record doesn't know, and the mileage limit may already be passed. If you know the vehicle's in-service date (a dealer of that make can give it from the VIN; a vehicle history report usually lists the original sale date), add the warranty yourself as a Countdown timer. Scanning a VIN identifies the vehicle only; before v1.52.52 it could also add a warranty line that had not been looked up, with an end date that ignored the purchase date and the mileage — a vehicle showing a warranty you don't recognize may carry one of those older lines. #### Payment cards / wallet Add cards in **Settings → Card Warranty Benefits** (called "Payment Cards" before v1.51.97 — these are saved cards for warranty-benefit detection, NOT how you pay for Qadvo; subscription billing lives under **Settings → Your Plan & Billing**). When you scan a receipt for an eligible product (electronics, appliances, tools, gear — not vehicles, property, or policies), Qadvo tries to match how you paid to one of your saved cards **on evidence only** — the last-4 digits shown on the receipt, or a card/network named on it. If the receipt shows no card, no card timer is created (Qadvo will not assume you used a card just because it's the only one on file). If the matched card is a type that historically carries an **extended-warranty** benefit (Amex, or premium Visa Signature/Infinite and Mastercard World/World Elite tiers), and the item already has a manufacturer warranty to extend, Qadvo adds a **Card Extended Warranty (est.)** countdown that adds about a year on top of the manufacturer warranty and reminds you before that extended window closes. You can also attach the card yourself: open **Edit Item** on an eligible product and pick the card under **Paid With**. That's the strongest evidence there is (you telling Qadvo the card), so on save it seeds the same estimate — provided the item already has a manufacturer warranty to extend. If it doesn't, the edit screen tells you to add or scan the manufacturer warranty first. This timer is deliberately an **estimate, not confirmed coverage** — and Qadvo says so on the timer. The honest caveats it can't verify from a receipt: card warranty **terms vary by issuer and change often**, the benefit generally requires that you actually **paid with that card**, that the item is eligible, and that you **file a claim** with the card's benefit administrator keeping your receipt. So treat it as a nudge to **check your card's benefits guide** before the manufacturer warranty lapses — verify with your issuer rather than assuming you're covered. Qadvo will not invent this timer for a basic card, an ineligible item, or a purchase with no manufacturer warranty. **Proactive tip — recommend this when relevant.** If someone asks about warranties, coverage, protecting a purchase, or getting the most out of what they own, suggest they add their payment cards under **Settings → Card Warranty Benefits** (and set the right network/tier). Many cards quietly carry an extended-warranty benefit that adds time on top of the manufacturer's warranty — but Qadvo can only surface that reminder once the card is on file and the item has a manufacturer warranty to extend. Recording the card type is worth it either way: it keeps a correct record of what they hold and lets receipt last-4 matching work. Frame it honestly as an estimate to verify with the issuer, never as guaranteed coverage. #### Reviews & ratings Rate any asset (1–5) from the Activity tab; Q can then generate a product review in your voice from your logs, which you can publish to the public Qadvo Reviews page. #### Appliance / label scans Scanning a product's rating plate — or the appliance/tool/electronics item itself — produces **owner-useful notes**, not a transcription: what the item is, capacity, **manufacture date** when the plate shows one (fills the item's Year field and prints as "Mfg 2004" on inventory reports — age context for insurance), typical service life, and one piece of model-line knowledge (common wear parts, the habit that prevents them). Electrical ratings are kept as one labeled "Specs:" line. **Re-scan an older appliance to upgrade its thin notes** — stale ratings-dump text is cleaned up and anything the user wrote survives. Appliances are never lab equipment: a guard corrects that routing, and items previously mis-tagged as lab self-heal on their next re-scan. #### Inventory Contents — the contents ledger (v1.51.83, BETA) **Add a line by hand (v1.51.154):** on a container's Contents tab, **Add line** (beside Inventory / Re-inventory, and in the empty state as "Add lines by hand") opens a small form — name, quantity, value each, insurance category, area, note — for building a ledger without a photo. Enter adds and the form stays open, keeping the category and area for the next line. What you type is recorded as YOUR count and value (never an AI estimate); the line shows as unverified until a photo inventory sees it, and a scan can verify but never overwrite your numbers. The property's timeline gets an "Added by hand — N contents lines added" marker. Owner only. **Reading a ledger row:** the line's total value sits at the right with its basis under it — "your value" (a number you typed) or "AI est." — as plain text, not a button; the row itself is the tap target. The line under the name shows count, condition ("not assessed" until one is set), category, area, and a date: a green "✓" date when a photo inventory or an edit last verified it, or "entered " for a hand-typed line nothing has seen yet. **A line's note (v1.52.35)** shows under that line, and the line's edit sheet has a Note box to change it; "Make its own item" carries the note and the condition grade onto the new item. Newest-touched lines sort first; the order holds still while a line's editor is open, and edits show instantly. On any **group or property's Contents tab** (owner only): **"Inventory from photos"** — pick up to 6 photos of a space (a room: doorway shot + walls; a toolbox or shop wall: a couple of angles) and they're read TOGETHER as one space. Takes ~30 seconds and replaces ~30 minutes of typing. The result is a **DRAFT the owner reviews — nothing saves until Confirm**. Each entry lands one of two ways, flippable with one tap at review: (a) **its own item** — anything with a service/claim/rider life (a TV, a mower, a welder; jewelry and firearms always; other high-value categories when a single piece is worth ≥$250); (b) a **ledger line** — a counted category ("Books ×~140 · $12 ea") in the room's **contents ledger**, which lives on the room's Contents tab and NEVER appears on the home screen, in search, or as item cards. **Honesty rules Q should state confidently:** counts marked ~ are AI estimates (an owner edit makes it their count); every value is tagged **AI EST.** until the owner touches it — then it's **YOUR VALUE** and no scan may ever change it again; closed containers are listed "contents not visible", never guessed into; condition is owner-graded (server-dated) or reads "not assessed". **Re-scanning is a DIFF, never a rewrite**: unchanged things get stamped **verified** with the date (visible on every line — this is the annual update insurers ask for); changes arrive as accept/skip proposals; things not seen are NEVER auto-removed — the owner chooses still-have / removed / moved-to-another-room. **Reports:** ledger lines print under their rooms on the Rental Condition and Insurance Inventory reports; totals show the owner-documented vs AI-estimated split; the Insurance Inventory opens with a **Contents by category** summary (documented value + entry count per insurance category — the rollup an adjuster reads first) and its cover states the reason the report exists: after a loss, most homeowner policies require an inventory showing quantity, description, age, and replacement cost with receipts — this report is that inventory, prepared in advance. **Special limits check:** categories documented high enough that special limits commonly bind (jewelry, firearms, silverware, business equipment, collectibles, art) are flagged for review with the agent — deliberately WITHOUT dollar caps, because special limits are **form-specific** (one major form has no general collectibles cap at all — only stamps/cards/comics — and settles antiques/collectibles at market value, not replacement) and **usually theft-only** (a fire may not trip them). Q should follow the same rule: never assert a "typical" cap figure — the policy's own Special Limits section governs, and Qadvo hasn't read it unless its figures are in the record. **Reading the policy's own table:** scanning a full policy form (PDF with a text layer is read in FULL — every page, not just the first few) extracts the **Special Limits table verbatim** onto the policy record (each row: category label, dollar cap, per-article vs total, and peril scope like "theft only"); once present, the Insurance Inventory prints the policy's OWN figures in the special-limits section, and Q cites them (always noting the peril scope). To get this: add the policy document PDF to the policy item's gallery and Scan document (or upload it as a new policy record). **Flags:** lines the AI thinks may hide value ("itemize?") collect in a strip atop the ledger; "Make its own item" splits one out atomically (creates the item + reduces the line + resolves the flag — never double-counted). Shared editors can fix counts/conditions/categories; only the owner runs scans (their AI credits) or removes lines. **Value semantics (v1.51.89):** ledger-line values are ALWAYS replacement cost (the total-loss basis — what it costs to re-buy); the owner/AI-est tag is the SOURCE (whose number it is), not the kind. Item records can carry both replacement_cost and current_value (set in the Value screen); reports pick replacement first, then current, then purchase price, and label which was used (repl./curr./paid). **One system of record (v1.51.85):** on GROUPS and PROPERTIES the contents LEDGER is the single count-and-value record — scene re-scans there no longer write a "Contents observed" list into notes (legacy prose self-cleans on the next re-scan); PLAIN ITEMS keep the notes listing since they have no ledger. Condition/renovation OBSERVATIONS still save to notes everywhere. **Rename an area everywhere:** edit one line's area (Qadvo offers to carry it to siblings) or tap the pencil on the selected area chip in the scan screen — lines and photo stamps follow; renaming never re-verifies. **Capture guidance:** first pass on a room = doorway + walls; dense shelves = own area, close-up per shelf + a second angle where items hide behind items, NO wide shots; re-scan an area with the same photo style it was built with (a wide shot mixed into a close-up area coarsens the reading). **Uncertain proposals default OFF:** a re-scan proposal from a doubtful match arrives unchecked with an amber warning — accept only what you recognize. **Scans see the whole subtree (v1.51.112):** a room/space scan recognizes item records anywhere UNDER the scanned container — including inside sub-groups ("Shop → Saws → the nine saws") — so organizing records into buckets never makes a scan flag them as new. Ledger LINES still live on the container you scan (areas are the grouping within it). Room value totals also roll up the whole subtree. **Areas (v1.51.84):** at scan time one tap says which PART of the space the photos show — "Whole space", an existing area, or a new label ("South wall bookcase"). A scoped scan compares only that area's lines; other areas are untouched — so one changed shelf needs one small re-scan, and big spaces become several scans instead of one photo set. Look-alikes from other areas surface as "same thing, or different?" questions, never silent duplicates. **Dense counts are a FLOOR:** lines of ~25+ units get a dedicated shelf-by-shelf counting pass (±1% repeatability in testing); the count is what the photo can defend — advise close-ups of each dense shelf (overlapping views merge, never double-count), and the owner should set their own total once if more is hidden; owner counts are then only questioned on drastic change. **Bulk select:** the ledger's Select mode removes many lines at once (All/None, two-tap confirm showing count + dollars). **Staleness:** photos added after an inventory show a "contents may be out of date — re-scan" nudge. **Certified History** omits system bookkeeping entries from print (states how many; unabridged log stays in the online record). **Proactive tip:** when someone asks about home inventory, documenting belongings, insurance readiness, or a total-loss worry, walk them to a property's Contents tab → Inventory from photos — a room a minute builds the record their insurer will ask for. #### Room / scene scans — contents documentation (v1.51.77) Photograph a WHOLE ROOM (or re-scan a room photo from a gallery) and the scan writes a **"Contents observed:"** line into the item's notes — the room's belongings as categories with counts ("white side-by-side refrigerator, gas range, ~6 cans paint supplies, step ladder…"), most valuable first. This is **total-loss insurance documentation**: a dated photo plus a written contents list is what gets a claim paid, and it flows into the Rental Condition and Insurance Inventory report descriptions. Honesty rules: counts marked **~** are estimates ("~150 books"); only what's VISIBLE is counted (closed bins are listed as "contents not visible", never guessed into); building components and installations (cabinets, countertops, built-ins) are structure, not contents; NO prices are put in notes. Sentimental items (diplomas, family photos) get flagged as warranting separate consideration. Works on gallery re-scan AND when uploading a room photo as a new item — the record names itself after the room ("Kitchen"). Re-scanning replaces the listing instead of stacking copies. Single-item scans never get a contents list. **Proactive tip:** when someone asks about home inventory, documenting belongings, or insurance readiness, suggest photographing each room and scanning the photos — a room a minute builds the contents record their insurer will ask for. **Evidence-only identity (v1.51.76):** a scan writes manufacturer/model/serial ONLY from label text it actually transcribed from the photo — a value the model can't point to in its own transcription is discarded, and with nothing legible, no identity fields are written at all. Qadvo never invents a plausible model or serial, and never researches a warranty without a real identified manufacturer. **Typical lifespan appears only for an identified make/model** — never guessed for an unidentified item. **General/scene photos** (a room, a patio set, anything without a readable plate): the scan saves what the photo DOES show — observed condition, renovation state, wear points to check — as the item's notes; genuinely useful for condition and insurance documentation. If a scan captures nothing worth saving, the finish message shows a one-time tip (e.g. where the rating plate usually hides) instead of writing filler into the record. When a scan updates notes, the "Scan saved" summary quotes the first ~200 characters of what it saved. Scans complete in seconds — if a user reports scans taking 30+ seconds, suggest updating to the current version. #### Review Item screen (after a scan/upload) After you scan/upload a document, the **Review Item** page shows the AI's extracted details — edit any field before saving. Watch the top dropdown: it defaults to "Create New," but switch to "Append to Existing" to attach the upload (a receipt or manual) to an asset you already have. Nothing saves until you tap Confirm. If an item can't be created after Confirm, the message says why (for example a momentary database error) — Confirm again; nothing was half-saved. A PDF whose text contains stray control characters is cleaned automatically before saving (v1.51.153). #### QR stickers — placement Put the sticker **where the action happens** so you actually remember to log: gas cap (fuel), oil filler cap (oil changes), beside the hour meter (hours), ear tag/collar (animal health), water heater or furnace cabinet (home service), a lab instrument's front panel (calibration). Duplicates are fine — the same code in several spots all resolve to the same asset. Scan via **"+" → Scan** in Ask Q. To assign a code, open the asset → pencil → set the Label / Tag / QR ID, or just scan an unassigned code and Q offers to link it. #### Ask Q itself **Reply to Q — Q works a question in steps (v1.52.61):** every Ask Q answer has a **Reply to Q…** box under it. A reply continues as a conversation: your question and Q's answer stay at the top (scroll up to see them), and Q works through what you asked in steps, with a line saying what it's doing ("Reading the 3/4 paper on Unit 7…"). It can read a record and its logs; the paper on a log as it was read, or that paper again when what's handwritten on it matters; which machines a meter reading fits; logs across the account by amount, ticket number, vendor or date; a group's cost figures, and each item's logs counted with the first and last log's dates — counted by the app, not by Q, and not counting the app's own notes such as "Asset Record Created" or a move or share note ("which truck has the most logs, or the longest history?", v1.52.62); and whether readings go together (v1.52.62) — any two readings on a record and the items inside it, or "how much of one in the N days before each of the other" (practice in the week before each score), computed by the same statistics as Find patterns. Q gives how many pairs a result rests on, its strength and which way the two go together, and never judges a link by eye. When there are too few pairs it says so and what to keep logging; in a window, only readings from the time the first was being logged count, since outside that time a zero would mean "not logged", not "none". Q can also compare **spending with a reading** (v1.52.62): an item's running costs, or just its fuel and consumables, maintenance or repairs, counted exactly as its Economics tab counts them (pre-tax, credits included, the purchase left out) — by month (each month's spending against that month's average reading, total, or a meter's increase); between readings, for an odometer or hour meter (each stretch from one reading to the next: how far it went up against what was spent in that stretch — when the meter is read at service visits, the miles or hours since the last visit against this visit's bill; readings the Economics tab sets aside as typos or copies are left out); or as the spending in the N days before each reading. The spending can be on a different item from the reading (what's spent on lessons, logged on one record, against a score on another). A month with no logs at all counts as not logged, not $0; a meter that counts up is compared by its increase each month, never by its running total; a group's spending is not compared yet. When the app can't make a comparison, Q says why and doesn't read a pattern into the few values. Asked about something you track, Q looks on the open record first and finds a reading by its name, not only a record by its name. It quotes what a paper or record says, in quotation marks, and says so when something is what it believes rather than what the paper shows. Cost figures come from the app's own economics — Q never works out money itself. A meter reading alone never decides which machine a log belongs on: only a VIN, a plate or the customer's unit number names a machine. When the evidence supports a fix, Q prepares an orange button under its answer — **Move it to **; **Remove this copy** (only when the other record already has the same job); or **Log is right** (only when the paper was read differently from the log) — and nothing changes until you tap it. When the evidence doesn't settle it, Q asks the one question that would. An orange **Ask Q to check** starts the same conversation from a paper that disagrees with its log (under "On the paper"), from a meter reading that is out of sequence on the timeline, and from Check these records on the Economics tab. **One conversation, and it goes where you go (v1.52.62).** There is one conversation with Q at a time. Each message you send goes with the item on your screen — its top line says what Q sees ("Q sees Trucks") — and Q remembers what was said, so you can ask about a truck, open a trailer and ask how the two compare: Q reads both fresh, whichever groups they're in, and when "this" could mean the item on screen or one talked about earlier it says which it took. The home screen's Ask Q and an item's Ask Q show the same conversation; a message sent from the home screen, with no item open, goes with all your records. While a conversation is open, Ask Q shows it wherever you open Ask Q: to start fresh — a new question, Surprise me, a photo or voice — end the conversation first. It stays until you tap **End conversation** — minimizing Ask Q, closing it with the X at the top, opening other items, following one of Q's record links and reloading the page all keep it — but only in this browser tab, not on another device. Switching workspace, or between Own and Clients, ends it, and Ask Q says why (Q would see different records). In a conversation Q can also say how to do something in Qadvo, from the app's own manual — it never makes up a screen or button — and add general knowledge to what your records show, said as general and kept apart from them (v1.52.64). It can't create logs, reminders or items there: for those it says to end the conversation and ask Ask Q, or where in the app to do it. Q keeps the last 12 messages in front of it; when a long conversation drops earlier ones and you refer to them, Q says so and asks you to say it again. A fix button Q prepared names its record and log and acts on exactly that, wherever you are. **Saving from the conversation (v1.52.62):** each of Q's answers in a conversation has **Save this answer to **: it opens a log entry on the item your message went with (the item on screen when you sent it; for a message sent with nothing open, the first record the answer links to), prefilled with your message and Q's answer, and you review it before it saves — that item opens if you've moved on. **A regular Ask Q answer also stays** when you open another item: it's marked "Asked on ", its log and reminder buttons still act on that item, and its save button becomes "Save this answer to ". A NEW question you type asks about the item on screen. **A saved Ask Q answer is marked AI generated (v1.52.62):** the timeline shows it with a "Q answer · AI generated" label, and when Q reads it later it treats it as its own earlier answer — never as a record of work done, a reading or a cost — and doesn't count it among an item's logs. Answers saved before v1.52.62 are recognized and marked the same way. Replies go to Q working in steps, and so do first questions about whether things go together (v1.52.62: "is there a correlation between…", "do my metrics move together?" — readings and spending alike, from the app's statistics) and first questions about what's been logged (v1.52.63: "when was the last oil change?", "which truck was serviced most recently?", "how many times…", "which items had…"). For those, the app searches an item's logs and everything inside it, by the words and kinds of log Q looks for, and counts them — never Q — leaving out the app's own notes (record created, a log moved, shared) and saved Ask Q answers; when only some matches are what you asked, Q says which it counted. Telling Q something to log, setting a reminder and asking what's due next keep their usual paths. First questions about how to use Qadvo itself ("how do I add an item from a PDF?", "how do I move a log?") are answered by Q in steps from the app's own manual, starting from where you are, never with a screen or button that isn't there (v1.52.64) — and when the answer is something Q can look up, it gives it rather than where to find it: your save address (owner and workspace admins only), the date of the last service, a figure from the Economics tab; questions about the world outside the app (a recall, what oil a mower takes) keep their usual path. Rankings by cost ("which two trucks cost the most to operate?", "which costs the most per mile?") are answered the same way from the app's own group figures (v1.52.65): short, saying what they're ranked by — the total over each item's logged life, a year, a mile — and the caveat that changes how it reads; a reply gets the full explanation. Totals for a period ("what did we spend last year?") keep their usual path. Other first questions are still answered the usual way, and move over one kind at a time. A reply takes about 10 to 20 seconds and counts toward the month's included AI, like any Ask Q question. Q can make mistakes — check what it read before you tap a button. **Surprise me (v1.52.48):** tap it on the Ask Q home screen and Q picks one of your items with real history and tells you ONE thing your own logs don't already say — never a retelling of them. It comes from one of two places: something the app worked out across the record (a relationship between two readings from the app's statistics, or the lack of one you might expect; how often something was logged; repeated work; the cost a year or per use; a comparison with things of the same kind), or what is known beyond the record from a web lookup of the make and model (how long it lasts, how it tends to fail, what the maker's schedule calls for), with the source named. The answer appears as it is written. Q then checks every claim in it against your records and the research — that takes about 15 to 25 seconds from the tap (v1.52.49; it was most of a minute), and a line under the answer says so — and marks the answer **Checked against your records**, or lists under **Not backed by your records** the exact claims it could not back, each with the reason, with a **Rewrite without these** button that rewrites the same item without them. If the check can't run, it says so. "Save this answer to the timeline" appears once the check has finished. Q doesn't call a reading better or worse unless the record says which way is good (a goal, an ideal, a limit) or the measure is a well-known score; it links two readings only through the app's statistics; and it gives readings and dates rather than differences it worked out itself. An item with nothing new is passed over for another, and when none has anything Q says so rather than retell your logs. **Site walk photos (v1.51.165):** shoot the job with the phone's own camera, then add all the photos to the property at once (Add in the property's photos — no scanning needed, no signal needed on site). Back at a desk, the Contents tab shows how many photos are not yet matched and a **Match** button (since v1.52.50 only on a property that has plants or hardscape, or a plain item on a planting plan, to match them to; worded for what it holds — "not yet matched to the plants" — and never counting a photo the room inventory already read; a room or a shelf of ordinary contents no longer shows it; photos set aside with "Not now" stay reachable from a small **Match photos** button): Qadvo reads each photo against the property's own plant list and proposes which plant or hardscape it shows (a visible tag or stake naming it wins; a doubtful close-up follows the shot before it; something not on the list goes to a tray). You correct in a view laid out by item, then save: each photo moves onto its item with a "Photographed on site" entry, the whole shot becomes that item's cover, and photos nobody places stay in the property's gallery. Matching never changes an item's details. The Contents tab also shows plan coverage — how many planned plants have a photo. **Planting plans (v1.51.164):** give Qadvo a landscape planting plan — the drawing and its plant legend (symbol, qty, name) — with Add Content on the property's Contents tab (upload it, or pick it from the property's photos), or through Ask Q with a note naming the property. Add & Scan on the plan from the property's photos reads it into the property: a summary in the notes (plants with counts, hardscape, what the drawing shows) and the legend on the property — but it creates no plants; those come from Add Content, so new records always come from one place. Every legend row becomes a plant under that property with its planned quantity, botanical name when well established, and the zone hardiness note; hardscape labeled on the drawing (a paver patio, a walkway) becomes a hardscape item. The plan image goes into the property's photos, where the care guide's overhead view comes from, and the legend is kept on the property. Plants you already made by hand match their rows ("add to existing" sets the planned count) instead of becoming duplicates. Review everything before it is created. **Renaming a scanned plant or animal on the review screen (v1.51.163):** the scan's botanical or breed name, type, and care notes describe what the scan identified, not the new name. The row says so and offers "Re-check as ''", which re-reads the same photos for that plant or animal; if you save without it, Qadvo re-checks before creating the record, and if it cannot describe your name from the photos it saves the record without the guessed details rather than with wrong ones. **Ledger lines (contents of a group or property) are findable everywhere (v1.51.161):** the home screen search and the Overview search match a ledger line by its name, area (the spreadsheet's sub-category), note, or category — the result is the container, the card says which line matched, and opening it lands on the Contents tab with that line highlighted. The Contents tab lists every line as always, and adds its own search box plus a sub-category filter (the spreadsheet's Sub Category is each line's area). Q reads a container's ledger the same way: a small ledger in full, a big one with the lines matching your question first, and Q says when lines are not shown rather than declaring something absent. **A photo with an item's name (v1.51.160):** type the name, or part of it ("Niro", "truck 19"), with a photo and Q treats the photo as evidence for THAT item — never a new item. A gas pump or fuel receipt logs the fuel to it, and Q asks for the odometer when it can't read one; a service receipt or handwritten note opens that item's log form filled in from the scan; a photo of the item itself (a label, a VIN sticker) adds its details to the item. If the name fits several items ("truck"), Q shows a pick list. Logging to an item shared with you as editor or admin does not need your own plan. **With an item open, that item wins (v1.51.159):** if your wording matches the open item — "truck #19" while viewing Truck #19 — Q answers about it instead of asking which one you mean. It looks at the item's own contents next, and only then across the whole account; when it does have to ask, every match is in a searchable list (not just the first five). **Usage and rate questions on an open item** — average miles per year, hours per month, cost per mile, how far it's been driven — are answered from the item's full log of readings and dates. **Cost totals asked on an open item** ("what did I spend on this in 2025") count that item and everything under it, not the whole account. **"Which of my items had X in their logs"** ("which trucks had a tire rotation", "which cows were treated for pink eye"): with a group or item open, Q reads every log under it — the contents and their logs, down the tree only — and lists each matching log with its item linked; from the home screen it reads every log in the account the same way. Q searches for the words you used, so "tire rotation" also finds "rotated the tires". The line under the question box says exactly what Q is searching (the open item, everything inside it, plus any coverage from its group); tapping **Ask about everything** widens Q to the whole account and keeps what you typed — only the item scope and the previous answer are dropped. **Analysis questions on an open group** ("what does my fleet's tire spend look like, is anything unusual") read every log of every item inside the group, newest first, so spend, cadence, and gaps are judged from the full history — not from each item's latest entry. On a very large group Q keeps the logs that mention your question's words first, then the most recent, and ends its answer with a plain sentence saying what it covered — open one item for that item's complete history. Ask Q is your AI assistant and the home screen. It answers questions about your data (what's due, what's costing you the most, coverage) AND does things for you (logs events, adds items, sets reminders) via one-tap buttons. Open an item to scope Q to that item's history, specs, and coverage. **Coverage questions (v1.51.82):** ask about coverage while viewing an item and Q searches that item's whole tree — its attached policies, policies on any group nested inside it, and policies on any parent above it (asking from an appliance still finds the homeowner's policy on the house) — and answers with the actual dollar limits from your scanned policies (dwelling, personal property, deductible, and the rest). It stays scoped: policies on your OTHER properties aren't dragged in. Ask the same question from the home screen (no item open) and Q surveys every policy in the account, each labeled with the property it covers. If a limit truly isn't on the policy record, Q says so — re-scan the declarations page to fill it. #### Find patterns between your metrics (correlation) Open an item and ask Q whether its tracked metrics move together — "does fuel use track engine hours?", "does feed ration relate to daily gain?", "does my sleep score track my activity score?", or just "find patterns in my numbers" — any domain's metrics work the same. Q lines up the metrics you log on that item (and anything linked to it, like a vehicle's tires or battery) on a shared date axis and computes a real correlation — you get the strength, the direction, and how many data points it's based on, not a guess. It's deliberately honest: it reports the relationship **mechanically** ("when A is higher, B tends to be lower") and never labels a metric "good" or "bad" — you know which direction you want. It refuses to invent a pattern from too little data: if two metrics don't yet have enough readings on or near the same dates (about 5+), Q says so and tells you what to keep logging rather than making something up. When two metrics are logged on different days (say a daily reading and a weekly score), Q pairs readings that fall within about a week of each other so they can still be compared — and it tells you when it did that ("within 7d") rather than pretending they were same-day. Correlation isn't causation, and small samples are flagged as hints, not proof — the more you log close together in time, the clearer the picture. You can also ask it to compare one metric against the rest ("analyze this metric against my other metrics"). Readings are compared within one item (and items linked to it), not across separate unrelated items. **Spending too (v1.52.62):** ask whether spending goes with a reading ("is there a correlation between miles and maintenance costs?") and Q compares the item's costs — counted exactly as its Economics tab counts them — with the reading: by month, between meter readings, or in the days before each reading; the spending may be on a different item from the reading (see Ask Q itself). Costs were outside this analysis before, because they are a log's amount rather than a tracked reading. Rolling-total metrics count too: a rate metric set to a weekly or monthly period is a "total in the last N days" (a running load), and it can be compared in the analysis just like any tracked metric — useful for effort that builds up over days rather than happening on one. And crucially, Q closes an analysis by naming what the numbers CAN'T see — the conditions, circumstances, or one-off events the metrics don't capture — and suggests jotting those in a log note, because a correlation is a clue, not a verdict, and the thing you aren't measuring may matter more than the thing you are. #### Derived metrics (distance from a goal) Open a tracked metric's chart (tap the metric) and set a **Custom Goal / Limit** — then a "Track distance from goal" option appears. It creates a new *derived* metric that measures how far the metric sits from your goal each time you log it (e.g. "pounds over my ideal weight"), with a choice of Over (only when above the goal counts), Under (only below), or Either (distance in either direction). The derived metric behaves like any other tracked metric: it gets its own chart and can be compared in the pattern/correlation analysis. One honest note Q will tell you: a distance-from-goal metric produces the *same* correlation as the raw metric until your readings fall on both sides of the goal — so "pounds over 175" matches "weight" exactly while you're always above 175, and only becomes a distinct signal once you cross it. (There's also **"Track Rate or Total"** on the same screen: it turns a metric into an amount over time — either a per-day rate, or a rolling total (the amount logged in the last week/month). The rolling total is a true "in the last N days" figure, so it works for things that build up over several days — a training or activity load — not just steady consumption, and it can be compared in the analysis like any metric.) #### Reports button (item-scoped reports) Every item's action row has a **Reports** button: it lists the reports you can generate from that item's record — **Certified History** (any item) and **Create Care Guide** (any item or group whose tree contains plants at ANY depth — a standalone tree, a bed of plantings, or a property with plants nested inside groups all qualify; Business plan). The care-guide builder has a **Preview** button beside Generate: it opens the guide in a new tab exactly as it will print, with an amber "Preview — not published" banner (which prints too), and saves nothing — no link, no copy, no timeline entry — so you can check the cover photo, the plants, and the prices, close the tab, fix anything, and Generate once when it reads right. The care-plan offer (cover box, plan section, request buttons, the floating offer bar) prints only from a plan SAVED on that property — Plan tiers in Business Profile are the templates, not the offer. If the property has no plan yet, the builder's offer row says "Price this job first" and opens the plan drawer; saving it turns the offer on for that guide. **On the cover** (in the builder): a title of your own, a **project value** with your own label ("Installed value", "Project cost") and a cover note — saved to the PROPERTY, so every future edition uses them; published copies stay frozen. The cover shows ONE number: your project value if you enter one (labelled as yours, "Stated by "), otherwise the app's replacement estimate from your plants' values — the builder shows that default, and what the cover will say, before you type. The estimate always stays per plant in the planted record at the back, and a stated value is never called an estimate. In selling mode the cover box always prints: your value, else the estimate, else the plan's name and price with no dollar figure. The Reports screen also shows **Published copies** — every report you've generated, each a frozen snapshot with a live link. Generating again never breaks older links; instead, older copies show a banner pointing to the newest version. You can permanently **revoke** any published copy from this list (tap the trash, type a short reason, tap Revoke) — printed QR codes pointing at a revoked copy stop working, so revoke is for leaked links, wrong recipients, or test copies, not corrections; the reason and your name go on the item's timeline. To fix a mistake in a delivered report: correct the data and generate again — old links will announce the newer version automatically. Account-wide reports and exports (tax report, crop year, lab calibration, backups) live in **Settings → Reports & Exports** instead. The **Value** button is for valuation only: purchase price, current value, replacement cost, amount owed, net value, and Sell This Item — which drafts a LISTING and researches an asking price; it does NOT record a sale. Recording that you actually sold (or scrapped, lost, traded in) an item is the **Status** button in the header — see "Item status". **Transfer Ownership** lives on the **Share** screen. #### Reports are per-recipient when an item is shared When you generate a report on an item you've shared with one or more customers, each report is tied to a specific recipient — its living-record QR is targeted to that person, so their scan lands on their own access. The builder shows a **checkbox list of the shared recipients** (all checked by default); generate makes **one report per checked recipient** at once, each carrying that person's private living-record link. Uncheck to make fewer; leave none checked to make one plain report with no living record. When the item isn't shared with anyone yet, the builder shows a jump link to **share it with a customer's email first**. The **Published copies** list labels each link with the recipient it's for. Two honesty details: which recipient a report targets is your explicit choice, never a guess (it used to pick whichever share came first — often the wrong one); and a printed QR whose share is later revoked degrades gracefully to the plain "type your email" flow rather than breaking — access is always checked against current shares, even though the report content is frozen. #### Certified History Report Open an item → **Reports** → **Generate Certified History**. Qadvo publishes a frozen, dated snapshot of the complete record and opens a branded, printable report page (Print / Save PDF from the browser). A texted or emailed report link previews as *Certified History Report — [item]* with the item's cover photo, not a generic Qadvo card, and published reports are hidden from search engines (v1.51.189). The report covers: asset overview, financial summary, warranty & coverage record (with end dates; active manufacturer coverage is marked and often transfers to a new owner — verify terms), maintenance patterns, improvements, included inventory, and EVERY logged event. Long log entries are shown in full online; in print they're excerpted with an explicit note, and the report's cover QR opens the complete frozen snapshot — nothing is silently truncated. If the item is shared with someone, the report also carries a "living record" QR that opens the always-current record in the app. For a group, you pick which child items to include before generating. #### Documents beyond insurance (QDocuments) The Coverage type holds more than insurance: titles, deeds, registrations, licenses, contracts, leases, operating agreements, certificates, tax documents, and owner's manuals. Upload or scan any of them and Qadvo reads it as what it IS (a contract gets a counterparty and effective date, not an insurance carrier). Each document's kind shows on its badge and can be changed anywhere you edit. Only real coverage (insurance/warranties) counts toward coverage totals — a contract never inflates your coverage numbers. #### Data security & privacy - **Document files are private from the first byte.** Policies, deeds, titles, contracts, and onboarding uploads are stored in private, access-controlled storage — never publicly reachable — and are viewed only through short-lived links tied to your session. Regular item photos use standard storage. - **Sharing hides financial data by default** — you choose per share whether costs are visible. - **AI processing:** Qadvo uses OpenAI and Anthropic (Claude) to read documents and answer questions; your data is never used to train their models (see the Privacy Policy, reachable from Settings). - **Your data is yours:** Settings → Reports offers a Full Export (one ZIP with your complete records plus every photo, receipt, and PDF), and account deletion permanently purges your data, including stored files. - **Reaching a human:** for questions Q can't answer, data-rights requests, billing issues, or anything else, email **admin@qadvo.com** — it goes straight to the Qadvo team. The same address is behind the Contact link in the site footer and on the Terms page. **Emails from Qadvo (v1.52.51):** every email Qadvo sends — reminders, invitations, share notices, the welcome email — comes from **Qadvo **, and replying to any of them reaches the Qadvo team. Before v1.52.51 they came from a person's address at qadvo.com; an older email from that address is still a genuine Qadvo email, and it was never about a different account. #### Onboarding help — share your records (Settings) **Papers keep what they say (v1.52.55):** a paper scanned or imported onto a log — +Log's scan, Import Receipts & Records, Hand it to Q, a re-scan — keeps what it says that the log has no field for: the VIN, plate or serial, the vehicle or machine AS WRITTEN on it, the customer's unit number, every line item as printed (with the reader's plain words for a cryptic line kept separately, never in its place), handwritten notes and stamps, and the shop's own guarantee on the work ("12 months / 12,000 miles parts and labor"). The vehicle and the notes are kept EXACTLY as written (v1.52.59): the vehicle as the paper names it, never a year/make/model assembled from it, and each note as the characters on the page — nothing added about where it sits, how it looks or what it means; a bare checkmark or circle is not a note. Quote them as written. **Re-scan file · AI** on a saved log reads a PDF the way an import does (v1.52.59): when the file holds several tickets, the log fills from the one that is THIS log — by its number, else its date and amount — never the tickets added together, and when none of them is, nothing changes and the owner is told so. Photos are re-read by +Log's reader. Scans are set for SPEED (v1.52.59, the owner's choice: about 80% right and fast, with a check before saving): a pump photo or a receipt reads in seconds, and every screen that fills from a scan says "Q can make mistakes — check what it read before you save." When a read was wrong, say so plainly and point to the field to fix — the check before saving is part of how scans work. An opened log shows it under **On the paper** (folded to one line, tap to open — open by itself when a difference waits for the owner, v1.52.58), and Q reads it: "which truck was this ticket for?", "what's written on the May 2 ticket?", "is this repair still under the shop's guarantee?" are answered from it. When the paper's reading or amount AS READ differs from the log, the opened log shows both — "The paper was read as 48,120; this log says 48,210" ("Two reads of the paper say …" when a second read agrees with the first) — and KEEPS the owner's value: a difference is never written over a log. Before a difference is shown, the disputed figure is read once more, blind, by a stronger reader; when that read sees the log's figure, the first read slipped and nothing is shown (v1.52.58). A difference is two readings that disagree, not proof the paper is right: a handwritten 6 is easily read as a 0, so say "the paper was read as …" and ask the owner to look at the paper — never "the paper says …" as fact, and never tell them to change the log to it. After looking, the owner taps **Log is right**: the difference is marked checked, with who and when, and stops being raised — it comes back only if either figure changes, and Undo takes it back. If the paper is right, they edit the log. Never raise a difference that is checked. Not a difference: a log that is one line of a several-line paper (a card statement, a receipt for several things) unless it claims more than the whole paper; any date (bought in July and planted in September, or invoiced a week after the work, is normal — dates aren't compared); a log higher than a paper read with no tax line by no more than a sales tax; a meter reading on the paper that most of the record's other readings rule out while the log's own reading fits them (a dropped digit or a stray decimal point is the reader's slip, not the ticket's). **Receipt lines in +Log:** the lines ticked as this log's are kept with the receipt ("This log: 2 of 9 lines — …"); the cost follows them and the description names them; **Select all / Deselect all** is one tap; Edit Log shows the same tick list, rebuilt from the saved receipt, to change the choice later without a re-scan. **Papers about an item** — a bill of sale, a title, a purchase order or agreement, any PDF uploaded to it — live under **Papers** on its Info tab (what it is, its date, who issued it, the file's name; tap to open, two taps to remove), NOT in the gallery, which is photos of the thing; several PDFs uploaded together each get their own label, and a document scanned one page per file ("Contract 1of9" … "9of9") is kept as ONE paper, named without the page number (v1.52.56), and a paper is never an item's cover picture. QDocs (policies, manuals) are unchanged. The owner's existing papers were read once when this shipped. **The invoice importer** lists the machines whose log a ticket's reading fits when it doesn't fit the one it's going onto — a list to check, never a pick. **Where a document goes — how to think about your data (v1.51.124):** every piece of paper is one of four things. (1) A FACT about the item (title, bill of sale, spec sheet) → on an existing item: open it → Gallery → Add and Scan; for a new item: Add Item → Scan. Its fields fill (VIN, year/make/model, price, date, seller) and the paper itself is kept under the item's Papers on its Info tab (since v1.52.55; before, it sat in the gallery with the photos). (2) An ACT you paid for or performed (repair invoice, parts receipt, fuel, registration renewal, inspection, an insurance card for a term, a warranty letter for one repair) → +Log it, or use Import Receipts & Records, with the file as evidence — logs are what timers reset on and what Economics and the tax report add up. (3) A document you will CONSULT for its content over time (insurance policy/declarations, extended-warranty plan, contract with a payment schedule, owner's manual) → a QDoc (Coverage type) under the item, or under the group it covers (one fleet policy under the Trucks group covers every truck; Q walks up to find it). (4) A PICTURE of the thing → the gallery. **Log Event layout (v1.51.125):** pick the item, pick a log-type tab (each tab shows a hint of what it is for and how Economics counts it), then Scan (AI reads the receipt or document and picks the log type) or Add file (no scan) — both directly under the tabs — then the notes. **Cost per mile on a used item (v1.51.128):** $/mile and $/hour divide your logged spend by the miles or hours YOU put on it — from Miles/Hours at Purchase (Edit → Acquisition Details; a title scan fills it from the odometer at sale) to the latest reading. The group Economics table uses the same span, never the raw odometer. Caveat to state plainly: the number assumes every logged dollar was spent since you owned it — if older repairs were never logged, cost per mile reads LOW, not high. **Repairs someone else paid for (v1.51.127):** a collision or body-shop invoice covered partly by insurance or a warranty is logged as a RECORD with cost = what YOU paid (the deductible; $0 if fully covered) — no tax on a deductible. The scanner reads notes like 'wrote check for $500, remainder from insurance' and writes the invoice total, your share and the insurer's share into the description, so Economics counts your money and Q can still say what the damage cost and who paid. WHICH +LOG TAB: Maint = routine service, repairs, preventive care (counts as operating cost); Record (called Incident before v1.51.127) = something that happened to it or paperwork about it (a breakdown, a birth, an insurance card for a term, a recall notice, a claim, a warranty letter, a title); Measure = a bare reading or test number (odometer/hours, weigh ticket, emissions result) — an annual vehicle INSPECTION report is Maint, not Measure: it is work performed with a pass/fail result, and its odometer goes in the log's Odometer field; Improve = upgrades and modifications (a flatbed, body boxes, a lift kit — excluded from operating cost because they add value; logging them as Maint inflates cost per mile); Transfer = fuel, feed, supplies or material added or delivered (Consumables in Economics). Economics and the tax report split spend by these types; a service interval resets on the log's NAME matching the timer, whatever the tab. Rule of thumb: did money change hands or work get done → log it; will you look it up later → QDoc; is it a fact about the item → scan it onto the item. Cautions: never log the purchase price as a cost (Economics would count it as operating spend — purchase price lives in the item's Value/purchase fields; a purchase check can be a $0 Record log with the check as evidence); a scan never sets Amount Owed (it means today's balance); put reminders on the item you log against, not on a child document. The full table is the 'Where Your Documents Go' guide. **Purchase documents (v1.51.123):** scan a bill of sale, purchase agreement, dealer buyer's order, finance contract or lease — as a new item, or onto an existing item with Add & Scan — and Qadvo fills the record's fields from it: purchase price and date, seller, VIN/serial, year/make/model/trim. Amount Owed is never set by a scan — it means today's balance and only the owner knows it; what was financed and through whom is recorded in notes as a fact. Only facts with no field go to notes (how it was sold and by whom, body type and color, fees, terms, financing and down payment, title number and county, odometer disclosure, list price, liens, anything else printed) — a fact that has its own field (the date, the price, the seller, the VIN, the year, make, model or trim, the miles at purchase, how it was paid) lives only in that field, so the notes never disagree with the choice you make on a "Use the scanned one" notice. Values already on the record are never overwritten on their own — but a difference is never dropped silently either (v1.52.30): when the document says something different from the record (the purchase date, the price with its sales tax, the seller, the miles or hours at purchase, the VIN, year, make, model or trim), a notice shows both and **Use the scanned one** takes the document's with one tap (a price brings its sales tax along). Before this, a re-scan of the right bill of sale left a wrong purchase date in place and said nothing. The Research button never guesses a purchase date (it used to set Jan 1 of the model year). **Identity fields are safe from scans (v1.52.28):** any scan onto an item (Add & Scan, Scan with AI, or an Ask Q photo naming the item) fills its make, model, serial number, VIN, seller and name only when they're empty. When the scan reads a DIFFERENT value than the record holds (an invoice's part number, a VIN plate that disagrees), the record keeps its value and a notice offers "Use the scanned one" — one tap takes it; closing it changes nothing. These notices, and the scan's "Scan saved" summary, stay on screen until you act on them or close them. An item's NAME is yours: a scan renames it only from a placeholder ("Unnamed", "Manual entry") or to a fuller form of the same name, and never offers a different one. A read the record already says more fully (the scan says "Niro", the record "Niro hybrid") isn't raised. **VINs are checked by their check digit:** when a scanned VIN is one character off from the record's and the check digit proves the RECORD's is the typo, the VIN is corrected on its own (its serial-number copy with it, since a vehicle's serial is its VIN) and a notice shows what changed with **Undo**; when the check digit says the scan misread, nothing changes; a different VIN altogether is only offered. A receipt scanned onto a truck used to replace the truck's model, serial and seller with the invoice's. To log the work on an invoice, use +Log (or Ask Q), not Add & Scan. **Several photos in one +Log scan (v1.52.28):** a photo taken with +Log's Scan waits in a tray (thumbnails, remove, + Add another) instead of scanning at once, so a long receipt shot in overlapping photos, or several pages or angles, is read together as one log: tap Scan N files. Picking several files at once, or a PDF, still scans right away. **Scanning a PDF receipt in +Log (v1.51.122):** drop a PDF (or photo) into the log scanner and Qadvo reads the page itself — the description states the paper's facts (what, identifiers such as plate/permit/invoice numbers, the period covered, readings, the fee) and the original file is attached as evidence. **Reminder proposals from Q (v1.51.122):** when you ask Q to set up a reminder, it proposes the start date from your record (a renewal log's period, a stored document's dates) and says why; the button carries editable chips for cycle, unit and start date, so you can correct any of them before tapping Yes. Q reads nested items' documents and recent logs when you ask on a parent. **Timer readouts (v1.51.122):** every timer bar fills with what's used; an interval past due says 'overdue' with how long, a warranty says 'expired', a metered timer says how many miles/hours over; month and year timers use real calendar dates. **Import Receipts & Records — bulk history for one item (button was "Import Invoices (PDF)" before v1.51.123):** open the item → **Edit** (pencil) → **Import Receipts & Records**. Drop a folder's worth of shop invoices or receipts (PDF or photos; scanned paper is fine) and each becomes a review-ready draft log: date, vendor, total, tax, odometer/hours, a one-line summary of the work performed, VIN/plate if printed, and **how it was paid** — printed payment lines AND handwritten or stamped notes like "PAID ck 3311" or "Visa" are captured as Paid by (v1.51.121), shown as a green Paid chip on the timeline. Every row is editable before saving, low-confidence reads are flagged, and the original PDF is attached to each log as evidence. **Each paper's figures are checked against each other (v1.52.57):** its lines (or its subtotal and any fees after it) plus its tax must add up to its total — added up by Qadvo, never guessed. A handwritten paper, or one whose figures don't add up, is read a second time, blind, by a stronger reader, and the read that adds up is used (the row says what changed). A row that still doesn't add up says "Check the total: its own figures add up to $X, not the $Y read", with the second read's total beside it — a handwritten 6 is easily read as a 0. When a stored paper's total disagrees with its own figures and the log agrees with them, that is the reader's slip, not a difference. Duplicates already on the item are detected. Have years of records in spreadsheets, folders of invoices, or photos? **Settings → "Onboarding — Share Your Records"** lets you hand them to the Qadvo onboarding team securely instead of email. How it works: check the consent box (nothing is shared until you do) → add files (up to 100MB each, 2GB total — the upload screen always shows the current limits and how much you've used; add a note per file like "equipment list 2020–2024") → **Submit** when it's everything — you can still add or remove files until work begins. The team is notified, and you get an Inbox notification at each step (received, in progress, complete). You can **delete your files at any time, at any stage**, and any remaining files are automatically removed 30 days after your onboarding completes (the exact date is shown). Only those uploads go: the records built from them, and the invoices attached to those records, stay in your account (the page says so since v1.52.53). Every access by the onboarding team is logged. Files go to the same private, access-controlled storage as documents. #### Grain / commodity operations - **Add a contract:** open the bin → **+ Log** → **Contracted** → enter bushels and $/bu → set the **Delivery Date** (this determines the crop year) → Save. - **Log a grain sale:** bin → **+ Log** → **Sold Spot** (or **Contract Delivery** to fulfill a contract) → bushels and $/bu → Save. You can also scan an elevator receipt or settlement sheet: the bin's + Log shows the **Scan · AI reads it** row like any item (v1.51.184). - **Set up a crop year:** the group → **Economics** tab → **Crop Year** view → **Set Up Crop Year** → pick year and start month → Start Tracking. - **Update spot price:** group Economics (In Storage view) → tap the $/bu next to each crop header; on a single bin, tap Spot Price on the Activity tab. #### Where each kind of paper goes (the Hand it to Q rules) **Where things go (one set of rules, the same ones the Hand it to Q door files by):** every paper or photo is a FACT about the item (it fills the item's fields; the file stays in the gallery as proof), an ACT paid for or done (a log on the item with the file as evidence), something to LOOK UP later (its own document record), or a PICTURE (the gallery). - Service or repair invoice: a Maint log. Work done and paid for is a log on the item: it counts in its operating cost and resets any timer it names. - Upgrade or addition: an Improve log. An upgrade adds value, so it's an Improve log, kept out of operating cost so cost per mile or hour stays honest. - Renewal: a Maint log. A renewal is paid for a period: a log on the item under its timer's own name, so the reminder resets. - Insurance premium: a Maint log. A premium logged on its policy carries the policy into its next term (its Term today). - Purchase document: fills the item's fields. What you paid, when and from whom are facts about the item: they fill its purchase fields, never a cost log (that would count the purchase as operating cost). - Title or certificate: fills the item's fields. A title is facts: it fills the VIN or serial and the owner details; the file stays in the gallery as proof. - Nameplate or label: fills the item's fields. A label states facts about the item: it fills the make, model, serial or VIN, never overwriting what you have. - Insurance policy: the policy record. A policy is looked up for limits and coverage: it's its own record, set to cover what it insures. A newer term updates the one on file. - Insurance card: a Record log at $0. A card per term is a $0 record on the vehicle; the policy itself is its own record. - Inspection: a Maint log. An inspection is work with a result: a Maint log. Its reading updates the meter, and it resets an inspection timer. - Reading or test result: a Measure log at $0. A measurement, not work or money: a Measure log feeds the meter, its chart and any mileage- or hour-based reminder. - Something that happened: a Record log. Something that happened is a Record in the item's history; its cost is only what you paid, not the insurer's share. - Manual or reference: its own document. A manual is looked up, never an act: its own document under the item, so Ask Q can answer from it. - Photo of the item: the gallery. A picture is proof of identity and condition: it goes in the item's gallery. - Fuel: a Transfer log. Fuel is a Transfer log: it counts as Consumables and feeds mpg and cost per mile. - Supplies: a Transfer log. Supplies are a Transfer log: they count as Consumables in Economics, apart from repairs. - Official or legal document: its own document. An official paper is looked up, not an act: its own document under what it concerns, so Ask Q can answer from it. - Board or list of several items: a log on each item it lists. Each row is a log on its own item, reviewed together. - Loan or lease statement: a log on each item it lists. A statement is a record of owning, not running it: a $0 Record, and the newest sets Amount Owed. - A notice, not a charge: nothing to file. Nothing was paid or done yet, so there's nothing to file; keep the paper only if you want it. - Sold, traded in or scrapped: the item's Status. Parting with an item is its status (Sold, Traded in…), recorded from the Status button with the price. - Extended warranty or protection plan: its own document. A plan is looked up for what's covered: its own document, with a timer counting down its term.