Set restaurant reservation turn times by measuring how long each physical table is occupied, then group those observations by party size and meal period. Choose a starting duration that covers a reasonably busy service, add any reset time your booking system needs, and check whether the resulting schedule fits your actual tables.
A single duration for every booking is simple, but it can leave tables idle at lunch or put dinner parties under pressure to leave. The goal is a workable assumption for each group, not a promise that every party will finish on schedule.
Measure from seating to a cleared table
For each party, start the clock when the guests are seated. Stop it when the table is cleared and ready for the next party to sit. That end point includes the time staff need to remove dishes, clean, and reset the table. If your staff record “cleared” before the table is ready, record the reset separately and add it before setting a booking duration.
Keep the unit of measurement clear. A party is a group of guests; a physical table is a piece of capacity. A party of four might occupy one four-seat table, while a party of six might use two tables joined together. Record both the party size and the physical tables occupied. If joined tables are unavailable until both are reset, use the time when the last one is ready as the end of that party’s occupancy.
For each observation, capture the meal period, scheduled arrival, actual seating time, ready-again time, party size, and table or tables used. A short note can explain an unusual delay, such as a celebration or a kitchen interruption. Keep guest details out of the worksheet unless you need them for a defined operational purpose.
Reservation turn time is different from arrival pacing. A pacing rule limits how many covers you accept in an interval. A turn-time rule estimates how long a particular table remains unavailable. Both affect the booking schedule, but a well-paced arrival list can still create a conflict if the assumed table duration is too short.
Check how your reservation system handles the start of a turn. If availability is calculated from the booked time rather than the actual seating time, a late-seated party may have less real table time than the schedule suggests. Decide how the host should handle that difference rather than quietly expecting guests to make it up.
Build durations for party size and meal period
Separate lunch from dinner and split observations into useful party-size groups. Start with groups your floor actually handles differently, such as parties of two, four, and six. Avoid creating so many categories that each has only a handful of observations. Weekends, special menus, and large events may also deserve separate review if they regularly change service.
The following worksheet is hypothetical. Each row contains five seated-to-ready observations, already including table reset. For this small illustration, the owner selects the fourth-shortest observation and adds a 10-minute scheduling cushion. The cushion is extra protection against normal variation; it is not another reset charge.
| Meal period | Party size | Observed minutes, shortest to longest | Starting duration |
|---|---|---|---|
| Lunch | 2 | 45, 50, 55, 60, 70 | 70 minutes |
| Lunch | 4 | 55, 60, 65, 70, 80 | 80 minutes |
| Dinner | 2 | 55, 60, 65, 70, 80 | 80 minutes |
| Dinner | 4 | 75, 80, 90, 95, 110 | 105 minutes |
For the dinner party of four, the selected observation is 95 minutes; adding the cushion produces 105 minutes. The 110-minute visit still exceeds that assumption. Staff need a plan for that possibility rather than treating 105 minutes as a hard departure deadline.
Five observations are enough to show the method, not enough to establish a reliable operating standard. Collect a broader set across ordinary services before relying on the result. Compare the suggested durations with what hosts see on the floor. If a category has little data, use a cautious temporary assumption and schedule a date to review it.
Before entering the numbers, ask the vendor whether durations can vary by party size and meal period, and whether the system adds its own reset buffer. If it does, avoid counting that buffer twice. Ask how the system handles joined tables and whether staff can adjust an individual reservation when they have a good reason.
Check the capacity cost of a longer turn
A longer assumption can reduce booking conflicts, but it can also remove an available seating. Model the trade-off on actual physical tables before changing every dinner slot.
Consider a separate hypothetical service with one four-seat table available from 5:00 p.m. The last permitted reservation start is 8:15 p.m. For this simple comparison, assume parties arrive on time, the table is ready at the end of each assumed duration, and there are no other holds or table combinations.
- At 90 minutes: starts at 5:00, 6:30, and 8:00 fit. That is three parties, or up to 12 covers if each party has four guests.
- At 105 minutes: starts at 5:00 and 6:45 fit. The next start would be 8:30, after the last permitted start. That is two parties, or up to eight covers.
The difference is one possible booking and up to four covers on that table. It is not a forecast of lost sales: those slots may not all be requested or filled. The example also does not calculate money. Fixed costs, variable costs, and exception costs are excluded, as are sales and payment fees.
Under these assumptions, use the longer duration if recent dinner observations and host experience indicate that 90 minutes regularly leaves too little time to reset. Accept the lower theoretical capacity as the cost of a more workable schedule. If the longer duration mostly creates idle time, investigate a shorter assumption or a different cushion before applying it broadly.
Repeat the check across your floor, including smaller tables and any combinations. A four-person booking cannot automatically use a table already reserved for another party, even if the total cover count looks manageable. Also check whether a late start would run past your kitchen or closing schedule; the last permitted start alone does not answer that question.
Handle exceptions and update the schedule
Review long and short visits with the host and manager. Ask what happened before removing an observation. A one-off power interruption may not belong in a normal-service assumption. A celebration pattern that occurs every weekend may belong in a separate rule or a host note. Keep the original record and note the reason for any exclusion so the next review is consistent.
Agree on how staff should handle predictable exceptions. A tasting menu, a large joined-table party, or a guest request for a longer visit may need an individual duration. An early departure does not necessarily make a table immediately bookable: staff still need time to reset it, and the next seating has to fit the rest of the floor plan.
Assign one person to approve changes, record the old and new assumptions, and tell the team when the change takes effect. For existing reservations, inspect the affected schedule before applying a new rule. Ask the vendor whether a change affects bookings already on the books or only future availability. Do not assume the system will resolve existing overlaps for you.
Before using a new setup in live service, run a simulated schedule check with sample parties and times. Confirm that the expected tables and start times appear, including a joined-table case if you use one. Keep the simulation isolated: it must not send guest messages or create real bookings, payments, or dispatches. Then monitor actual service separately, with a manager ready to adjust individual bookings when needed.
Revisit assumptions after a meaningful menu, staffing, or service change, and on a regular schedule during stable periods. Compare observed occupancy with the scheduled duration and note both recurring conflicts and unused gaps. Your restaurant management software may help organize the operating workflow; ask which reservation and floor details staff can review. If you use a restaurant POS system, ask whether its timestamps are useful alongside host records, and check that they match the seated-to-ready definition before using them.
Choose one meal period and party size to review first. Bring the observations, floor plan, and proposed duration to your next vendor conversation, and compare the reservations workflow against your operating requirements with Fleksa’s restaurant reservation system.
Frequently asked questions
Should every party of four get the same turn time?
Not necessarily. Lunch and dinner can have different patterns, and joined tables or special menus may need different handling. Start with broad categories, then split them only when the floor and observations support the distinction.
Should the turn time include cleaning and reset?
The availability assumption should cover the period until the table is ready for the next party. If the system adds a separate reset buffer, check its settings so that you do not add the same time twice.
What should a host do when a party stays longer than expected?
Check the next reservation and available physical tables, then choose an appropriate adjustment. Track recurring overruns for review; a scheduled duration is an estimate, not proof that a table will be ready at an exact minute.
