Controlled overbooking is the practice of accepting more bookings than there are seats, anticipating that a percentage of guests won't turn up (no-shows). Calculated well, it maximises occupancy without creating conflicts.
Overbooking is a capacity management strategy that originated in the airline and hotel industries. It consists of accepting more bookings than there are physical seats available, anticipating that a percentage of guests won't turn up (no-shows) or will cancel at the last minute. In restaurants, where the no-show rate can reach 15–20% in some venues, controlled overbooking makes it possible to maximise actual occupancy by compensating for expected absences. The key word is "controlled": it is not about accepting unlimited bookings, but about calculating precisely how many extra bookings you can take based on historical no-show data, minimising the risk of every guest turning up at once.
The basic calculation is simple: if you have 50 seats and your historical no-show rate is 12%, statistically you can expect 6 out of every 50 bookings not to turn up, so you could accept up to 56 bookings with a reasonable chance of not exceeding capacity. However, overbooking carries significant risks: when everyone does turn up (the worst case), you have to deal with guests who have a confirmed booking but no table. That causes extreme dissatisfaction, bad reviews and reputational damage that can far outweigh the benefit of the extra tables filled on normal days. That is why overbooking must come with clear conflict-handling protocols: predefined compensation (complimentary drinks, a discount on the next visit, priority booking), maximum acceptable waiting times (15–20 minutes with a complimentary aperitif) and alternatives ready to go (agreements with nearby restaurants to send guests there, with compensation).
The most sophisticated restaurants vary overbooking by day and time slot: more aggressive on high no-show days (Friday night), conservative on low no-show days (family Sundays, when no-shows are rare). Overbooking can also be selective: applied only to last-minute bookings (which have a higher no-show rate) but not to bookings confirmed well in advance or secured with a deposit.
Maximum bookings = Capacity / (1 – Expected no-show rate)
The basic overbooking formula divides the restaurant's physical capacity by the expected attendance factor (1 minus the no-show rate). If your restaurant has 60 seats and your historical no-show rate is 15% (0.15), maximum bookings would be 60 / (1 – 0.15) = 60 / 0.85 = 70.6, rounded down to 70 bookings. This means you can accept up to 70 bookings expecting, statistically, only 60 to turn up (70 × 0.85 = 59.5).
However, this formula assumes no-shows are evenly distributed, which rarely happens. A more conservative approach takes no-show variability into account: if your average no-show rate is 15% but varies between 10% and 22% depending on the day, you should use the lowest value (10%) to calculate the maximum safe overbooking: 60 / 0.90 = 67 bookings. A more sophisticated approach calculates the probability of excess demand using statistical distributions (binomial or Poisson) and sets overbooking at the level where the probability of a conflict is below an acceptable threshold (for example, less than a 5% chance that more guests turn up than there are seats).
Your restaurant has 45 seats and you analyse the last 6 months of data. The average no-show rate is 14%, but it varies significantly: 18% on Fridays, 12% on Saturdays and only 6% on Sundays. You decide to apply differentiated overbooking: Friday (18% no-shows): effective capacity = 45 / 0.82 = 55 bookings. You accept up to 54 (safety margin).
Saturday (12% no-shows): effective capacity = 45 / 0.88 = 51 bookings. You accept up to 50. Sunday (6% no-shows): effective capacity = 45 / 0.94 = 48 bookings. You accept up to 47 (very conservative, because no-shows are low and the risk of conflict is high).
You put a handling protocol in place: if there's a queue because of overbooking, you offer a glass of cava and canapés at the bar (cost €6 per person) while a table is freed up. If the wait goes beyond 20 minutes, you offer 20% off dinner. If you can't seat them, you call your "partner" restaurant 200 metres away, pay for their taxi and give them a complimentary dinner at your place another time. After 3 months of controlled overbooking: average occupancy rises from 82% to 94%, conflicts occur in only 2 of the 36 Friday/Saturday services (5.5%), and weekend revenue rises by 8%.
Controlled overbooking matters because it tackles one of the biggest profitability problems in restaurants head-on: tables left empty by no-shows. A 50-seat restaurant with a 15% no-show rate has an average of 7–8 empty seats every service that could have been filled. With an average spend of €35, that is €245–280 of lost revenue per service, more than €8,000 a month. Overbooking lets you recover part of that loss by accepting extra bookings that will statistically fill the tables no-shows leave empty.
However, badly managed overbooking is one of the worst experiences a guest can have: arriving with a confirmed booking and being told "there's no table" causes extreme frustration, one-star reviews and reputational damage on social media. That is why overbooking only works when three conditions are met: first, calculations based on real historical no-show data (not intuition), updated regularly; second, conservative overbooking that minimises the chance of a conflict (better to lose a little occupancy than to cause frequent conflicts); third, a conflict-handling protocol with generous compensation that turns a bad experience into a loyalty opportunity ("yes, there was a problem, but look how they handled it"). Restaurants that implement controlled overbooking properly increase their actual occupancy by 5–12% without a significant rise in complaints. Those that get it wrong end up abandoning it after a run of conflicts and bad reviews.
The difference lies in the data, the calculation and the protocol.
Zindra analyses your historical no-show rate by day, time slot and booking channel, and automatically calculates the optimal level of overbooking for each service. The booking system can be set to accept bookings automatically up to the calculated limit and alerts you as you approach the risk threshold.
Tools and content to go deeper into this concept.
RevPASH (Revenue per Available Seat Hour) measures the revenue generated by each available seat per hour. It is the most complete indicator of a restaurant's operational efficiency and real profitability.
The occupancy rate measures the percentage of available seats actually filled during a service. It is a key indicator of a restaurant's efficiency and the basis for working out its revenue potential.
Yield management is the strategy of adjusting prices and availability according to expected demand in order to maximise the restaurant's total revenue.
Table turnover measures how many times each table is occupied during a service. It is a key operational efficiency indicator which, combined with average spend, determines the restaurant's revenue potential.
A no-show is when a guest with a booking neither turns up nor lets you know. No-shows affect 10–20% of bookings in Spain and cost restaurants thousands of euros a year in empty tables that could have been re-let.
Every hospitality term with formulas, examples and benchmarks in a handy PDF.
Try Zindra free and keep food cost, staff, inventory and much more under control from a single platform.