The No-Show Risk tool lives in the Re-engagement hub, under the "No-shows" tab (any old link to the standalone No-Show Risk page redirects there automatically now). It doesn't ask you for any input beyond clicking Run — it scans your shop's own upcoming scheduled appointments and returns, for each one, a risk score out of 100, a low/medium/high risk band, a few plain-language reasons behind that score, and — for the riskier jobs — a short suggested confirmation/reschedule text. Nothing is booked, sent, or changed on its own: the drafted text sits on the page next to a Copy button, and you decide whether and how to actually reach the customer.
Running a scan
Open the No-shows tab and click Run. The scan covers appointments scheduled for the next several days out — the page says up front how far ahead it's looking. Scoring itself runs entirely off your own stored data (how far in advance the job was booked, whether a deposit is on file, whether the customer is new, their history of past no-shows and cancellations), so it happens the same way whether or not your workspace has live AI turned on. For appointments that come back at medium or high risk, the tool then hands those specifics to the AI, which turns each one into a short, friendly text-message draft: a plain confirmation of the appointment with a low-key invitation to reschedule if the time doesn't work anymore, with no pricing or links included. If a customer was already reached recently through the app's own messaging, their appointment is skipped on the next run instead of getting a duplicate draft.
Reading the results
Each scanned appointment shows up as a card with the customer's name, the appointment time, and a risk score/band badge — low in green tones, medium in amber, high in rose. Underneath sit a handful of small tags explaining what moved the score, in plain language: things like how far out the job was booked, whether a deposit was collected, whether the customer has no completed-job history yet, or how many past no-shows or cancellations they have on file. Jobs flagged medium or high risk also get a "Suggested confirmation text" box with a Copy button — there's no send action on this tab, so you copy the draft and paste it into whatever channel you actually text customers from.
The same underlying risk score also surfaces as a small badge directly on appointment cards on the Calendar page, so you can spot a risky booking at a glance without running the AI tool at all — that badge comes from the same scoring, computed for free with no AI call behind it.
Things to know
- Plan requirement: No-Show Risk requires the Growth plan or above.
- Stub mode: if the workspace doesn't have `ANTHROPIC_API_KEY` configured, a banner says so, and while the risk scores and signals are still computed live from your real data, the drafted messages come back as one generic, canned confirmation text instead of copy tailored to each job's specific signals — a placeholder for testing, not a real read of the appointment.
- Running a scan that drafts messages counts against the shop's monthly AI usage allowance; the small risk badge on the Calendar page does not, since that variant never calls the AI.
- It's read-only end to end — nothing it does creates, updates, or deletes an appointment, customer, or payment record.
- If nothing on your calendar is currently at elevated risk, it says so plainly rather than forcing a result, and it won't spend AI budget drafting messages when there's nothing risky to draft for.
- Only appointments still on the books with an active, upcoming status get scored — ones waiting on a deposit, or already completed, canceled, or marked as a no-show, aren't included.
Frequently asked questions
Q: Does the tool text customers automatically?
A: No. Running it never sends anything on its own. For appointments at medium or high risk, it shows you a suggested confirmation/reschedule text with a Copy button, and you paste that into whatever channel you actually use to reach the customer.
Q: What plan do I need to use this?
A: Growth or above.
Q: What actually drives an appointment's risk score?
A: Signals from your own data — how far in advance the job was booked, whether a deposit was collected, whether the customer is new with no completed-job history, and their record of past no-shows or cancellations. Each score comes with a few of these shown right on the card so you can see why it landed where it did.
Q: Why do the draft messages all look the same?
A: That means the workspace is in stub mode — ANTHROPIC_API_KEY isn't configured. The risk scores themselves are still real, computed from your data, but the drafted text is a generic, canned placeholder — not written for that job's specific signals — until a live key is set.
Q: Can I change how far ahead it scans, or which risk level gets a draft?
A: Not from this tab today — it's just a Run button, with the scan window and draft threshold set by the tool itself rather than exposed as fields on the page.
Q: Does checking an appointment's risk on the Calendar page cost anything?
A: No. The small risk badge on the Calendar page comes from the same underlying scoring but never calls the AI, so viewing it doesn't use any of the shop's AI usage allowance — only running a full scan from the No-shows tab, which drafts messages, does.