To tell whether canceled tables leave enough time to resell, measure the hours between each cancellation and its scheduled arrival. Group cancellations by notice window and service period, then compare those windows with bookings made for the same service. The comparison shows where replacement demand may exist; it does not prove that a particular new party took a canceled table.
Start with a few weeks of comparable services and check the result against your actual floor plan. A Friday dinner cancellation may be easier to fill than a Tuesday lunch cancellation with the same amount of notice. Party size, available seating and the time a booking arrives all matter.
Record the three times that matter
For each canceled reservation, keep the original booking time, the cancellation time and the scheduled arrival time. Use the restaurant's local date and time consistently. Cancellation lead time is scheduled arrival minus cancellation time. Booking age at cancellation is cancellation time minus booking time; it answers a different question.
For example, in a hypothetical service, a party books on Monday at 6 p.m. for Friday at 7 p.m., then cancels Friday at 1 p.m. The booking was held for 91 hours before cancellation, but the restaurant received only six hours of notice. Use six hours when judging the chance to resell.
Keep the scheduled service period, party size, reservation status and booking channel alongside those times. A channel might be the restaurant's website, phone or a discovery service. If a guest changes an arrival time or party size before canceling, decide which version you will use and apply that rule consistently. The latest scheduled arrival before cancellation is a useful starting point because it reflects the slot that became available.
Ask your reservation provider whether its history or export includes cancellation timestamps, changes, channels and local time zones. Also ask how it records staff-entered phone bookings and cancellations. A current reservation list alone may not retain the history needed for this analysis.
Do not mix cancellations with no-shows. A cancellation releases a booking before arrival; a no-show may leave the table held until staff decide the party is not coming. The guide to measuring no-show costs covers that separate problem.
Group notice by service period
Choose notice bands that match how your team works. More than 24 hours gives you at least a day to react. Four to 24 hours covers the day before or much of the service day. Less than four hours captures changes that may arrive while the team is preparing or already serving. Use your own boundaries if your booking patterns call for them, but write the boundaries down so every reservation lands in one band.
The following worksheet uses entirely hypothetical Friday dinner data over four weeks. Each row counts canceled guest parties, not physical tables. It also counts new parties booked for the same Friday dinner services during the corresponding time window before their own scheduled arrivals. A party is counted once, in the window when it booked.
| Notice window | Canceled parties | New parties booked in window | What to investigate |
|---|---|---|---|
| More than 24 hours | 4 | 3 | Check whether later demand fits the released arrival times and party sizes. |
| 4 to 24 hours, including 24 hours | 5 | 2 | Look at service-day requests and how staff offer available slots. |
| Less than 4 hours | 3 | 0 | Check walk-in demand and whether short-notice availability is visible. |
These rows show 12 canceled parties and five new parties across the three windows. They do not show seven lost tables, a resale rate or a count of unfilled seats. New parties could have booked space that was already open. Some canceled slots could also have been used by walk-ins or by bookings made in a later window.
Run the same view separately for other services rather than blending Friday dinner with a quiet weekday lunch. A small sample can be misleading: one large party or an unusual event can change the picture. Compare several similar weeks and note holidays, closures or unusual capacity changes before drawing a conclusion.
Your restaurant management software may help bring operational records together, but ask which timestamps and historical statuses you can actually view or export. If you use a spreadsheet, protect guest information and include only the fields needed for the decision.
Check whether new demand could fit the released space
The worksheet is a starting point, not a matching engine. Look at individual services next. For each cancellation, ask whether a later booking or walk-in could have used the space at a compatible time and party size. Check the floor plan and actual seating records rather than assuming that one party always equals one table.
Physical capacity limits the answer. A party of six may need two tables combined, while two parties of two may require two separate tables. A cancellation does not necessarily release a usable table if the tables are needed for adjacent bookings, the section is closed or the available time is too short for another sitting. Conversely, staff may rearrange the room and accommodate a party without using the exact tables originally assigned.
Within the hypothetical Friday sample, the two bookings made in the four-to-24-hour window are a reason to inspect service-day demand, not proof that two of the five cancellations were replaced. A new booking could have arrived before a cancellation, requested a different time, or fit an already-open slot. Even when the timing fits, you need seating information to know whether the room could accommodate it.
For a more useful manual check, review each service in time order:
- Note when the cancellation occurred, its arrival time and party size.
- Check what capacity was available immediately afterward, including existing bookings and any closed sections.
- Review later bookings and walk-ins for compatible arrival times and party sizes.
- Mark a possible fit separately from a confirmed seating outcome. If the records cannot establish what happened, leave the outcome unknown.
Check the demand you cannot see, too. A full online book may discourage requests, and an unanswered phone inquiry will not appear as a reservation. Ask whether online availability updates when staff cancel a booking, whether staff can adjust the released capacity, and how quickly a change appears to guests. The answers depend on the system and your setup.
Channel matters when comparing tools. OpenTable offers reservations, operations tools and a diner discovery network. A direct booking tool takes reservations through a restaurant's own website; Fleksa also offers Google reservation paths where available. Ask each provider about permissions, exports, channel attribution and fees so you can interpret demand and compare the workflow on the same terms. See Fleksa's reservation FAQ for more on its reservation offering.
Choose one policy test and define its limits
In the hypothetical worksheet, the four-to-24-hour band has the most cancellations, while some new bookings arrive during that window. A reasonable first test is to make service-day availability easier for staff to review and offer, rather than immediately changing a cancellation policy. The sample gives a reason to investigate that window; it does not establish that a stricter rule would improve results.
For the next comparable set of Friday dinners, assign one manager to review cancellations as they arrive and check whether the released capacity is accurately available to book. Record the notice band, party size, whether staff identified a compatible later party, and whether that party was actually seated. Keep unknown outcomes separate. Compare the same measures with earlier similar Fridays, while noting differences in staffing, events and open capacity.
Before live service, walk through a simulated cancellation in a practice environment or with a paper floor plan. Check who updates availability and who checks the result. The simulation should not send a guest message, take a real card payment or dispatch a courier. A live-service pilot is different: use genuine reservations under your normal operating procedures and tell staff who owns each action.
If you later consider deposits, cancellation deadlines or fees, ask the provider what the system supports and how staff would handle exceptions. Compare written quotes for any setup, subscription, payment processing and exception-handling costs. A processing fee alone is not the total cost of operating the policy. Discuss contract, legal or tax treatment with a qualified adviser where needed.
A useful result may be that late cancellations rarely coincide with compatible demand, or that the restaurant lacks enough data to tell. Either way, avoid treating a short sample as a forecast. Revisit the bands and the workflow after you have enough comparable services to see a pattern.
Next step: Pick one recurring service, collect its cancellation timestamps and review the notice bands with your floor manager. Then compare the workflow and current availability with Fleksa's restaurant reservation system.
Frequently asked questions
How is cancellation lead time different from booking age?
Cancellation lead time runs from cancellation to scheduled arrival. Booking age runs from the original booking to cancellation. The first tells you how long the restaurant has to respond; the second tells you how long the booking was held.
Does a new reservation prove a canceled table was resold?
No. The new party may have used capacity that was already available or needed a different time or table arrangement. Compare timestamps, party sizes and seating records before calling it a possible fit.
How many weeks of data do I need?
There is no universal number. Start with several comparable services, then extend the review if cancellations are rare or unusual events distort the sample. Treat uncertain outcomes as unknown rather than forcing a resale count.
