finnUse cases
Fill a canceled appointment from your waitlist.
Use AI to coordinate appointment cancellations and waitlists: match service requirements, manage offers, recheck availability, and confirm the booking.
A two-hour salon appointment is canceled for tomorrow afternoon. Three people on the waitlist would like an earlier visit, but they need different services and practitioners. The job is to find an eligible match and record a valid booking. This illustrative journey shows how Finn can handle the coordination while the business keeps control of its scheduling rules.
Confirm what is actually available.
Read the current booking record and verify that the cancellation is final. Check the available duration, practitioner, location, equipment, and any preparation or cleanup time. A two-hour gap does not necessarily support every two-hour service. Staff schedules and resource availability can change independently.
Keep the opening linked to a specific record. If the canceled appointment is restored or the practitioner becomes unavailable, the offer process needs to respond to that change. Do not keep sending messages based on an old snapshot of the calendar.
Match the waitlist using explicit rules.
Select customers whose requested service, practitioner preferences, location, and availability fit the opening. Apply the business’s queue policy, such as order of joining or an agreed priority rule. Avoid inventing a ranking based on unrelated personal characteristics or assumptions about who might spend more.
Check contact preferences and whether the person still wants an earlier appointment. Someone who already accepted another opening should not keep receiving offers. A customer’s existing booking also needs attention: replacing it and adding another appointment are different actions.
| Check | Question to answer |
|---|---|
| Service fit | Does the opening allow the requested service and preparation time? |
| Staff and resources | Are the practitioner, room, and equipment available? |
| Customer fit | Do the time, location, and stated preferences match? |
| Queue policy | Who should receive the offer under the business’s rules? |
| Contact permission | May this channel be used for this offer? |
Make an offer with a clear boundary.
State the service, practitioner, location, time, any price difference, and how the customer can accept. If the business uses an expiry period, show it clearly. Do not describe an offer as a confirmed reservation unless the booking system has actually created a hold or booking with the intended conditions.
Sequential offers are often easier to reason about: one person has the opportunity until they decline or the offer expires. If the business chooses concurrent offers, the system needs a reliable way to determine which acceptance obtains the slot. The customer-facing message must accurately explain those conditions.
Recheck the slot when the customer accepts.
Read current availability again and verify that the offer still applies. Another staff member may have booked the opening, or the customer may reply after expiry. Only confirm after the correct appointment is saved in the booking system. If the slot is unavailable, explain the change and offer the next appropriate option.
When moving an existing booking, follow the business’s change rules and avoid canceling the original before the replacement is secured. Check any deposit or payment requirement separately. A customer’s “yes” establishes intent; it does not prove the calendar or payment system has completed the work.
Stop the other offers and update the team.
Mark the opening as filled, close or withdraw outstanding offers as appropriate, and update the waitlist record. Send the customer confirmation and the relevant team update. If either message fails, leave that delivery problem visible with an owner. The business should be able to see which booking filled the original gap.
Anthropic’s evaluation guidance distinguishes an agent’s account of success from the actual resulting state. Apply that principle here by checking the saved appointment. In Finn, recurring or event-driven Tasks can connect the opening, customer exchange, and team follow-through under the agreed scheduling rules.
Measure filled slots and scheduling mistakes.
Track eligible openings, offers sent, accepted offers, valid bookings, time to fill, and employee effort. Review conflicting offers, expired acceptances, and changes that required correction. A high message count is not the outcome. A slot is filled only when the eligible appointment exists in the authoritative calendar.
Test simultaneous replies, a booking made by reception during the exchange, a disconnected calendar, an expired offer, and a customer who changes their mind. Start with one service or location so the team can inspect the process before extending it.
Common questions
Does offering an appointment reserve it?
Only if the booking system has created a real hold under the business’s rules. Otherwise the message should describe an offer subject to current availability.
How do you avoid double booking from a waitlist?
Use the booking system’s availability and reservation controls, recheck on acceptance, and confirm only after the appointment is saved. Stop or withdraw other offers when filled.
Is this the same as an appointment reminder?
No. A reminder concerns an existing booking. A waitlist workflow matches an opening to an eligible customer and creates or changes a booking.