Ticket times: the only kitchen metric that moves

Ticket times measure how long an order takes from the moment it reaches the kitchen to the moment it leaves the pass. They are the kitchen's speed, expressed in minutes, and they are one of the few kitchen numbers that respond directly to what the team does. Change the prep, the station layout, the menu or the pacing of the dining room, and ticket times move within days.

Many kitchens measure them badly or not at all. On paper tickets, the time an order arrived may be printed, but the time it left the pass rarely is. Where ticket times are measured, they are often reported as a single average for the night, which looks reassuring and hides the moments that matter: the Friday at half past eight when the grill fell twenty minutes behind and three tables waited for their mains.

This guide explains how to measure ticket times properly: from order to pass, by dish, by station and by hour. It covers the difference between ticket times and cook times, why the average hides the worst moments, how to find kitchen bottlenecks, what all day totals tell the line, how ticket times relate to speed of service in the dining room, and how to turn the numbers into changes that make the next Friday better.

A ticket time is the time between an order arriving in the kitchen and the finished dishes leaving the pass. If a table's mains are sent to the kitchen at 8:12 and leave the pass at 8:31, that ticket took 19 minutes. Measured across a service, ticket times show how quickly the kitchen turns orders into plates.

They are different from the time a guest waits for food, which also includes the time to take the order, the time the order spends before it is sent, and the time to carry plates to the table. Ticket times isolate the kitchen's part of the wait, which is what the kitchen can control.

They are also different from cook times, which measure how long a dish takes to prepare once a cook starts it. A ticket can wait several minutes before anyone starts it, especially at busy times. The difference between ticket time and cook time is queue time: the time an order spends waiting for a cook to be free.

Splitting ticket times into cook time and queue time is one of the most useful things a kitchen can do. Cook time tells you how long the dish takes when someone is working on it. Queue time tells you how long it waited before anyone did.

If cook times are steady but ticket times rise at busy moments, the problem is queue time: too many tickets arriving at once for the station to keep up. The answer is pacing, prep or another pair of hands at the bottleneck, not a faster recipe. If cook times themselves rise at busy moments, cooks are being interrupted, stations are short of prepared components or the dish is harder than it needs to be.

Kitchen display systems that record when each ticket arrives, when it is started and when it is marked ready can measure both. With paper tickets, it is much harder, though a few manual timings on busy nights can still reveal the pattern.

To measure ticket times, you need two moments for every ticket: when it arrived in the kitchen and when it left the pass. A kitchen display records both automatically as tickets are bumped. Without one, a manager with a notebook on the pass can time a sample of tickets on a few busy services.

Be consistent about what you measure. The arrival time should be when the order reached the kitchen, not when it was taken at the table. The finish time should be when the dishes left the pass, or when the ticket was marked ready, used the same way every time. Held courses need care: a main that was held on purpose until the starters were cleared should be timed from when it was fired, not from when it was ordered.

Record the station, the dishes and the time of day for every ticket. Without them, you have a list of times; with them, you can see where and when the kitchen slows down.

The most common way to report ticket times is a single average for the service: "average ticket time 14 minutes". It is easy to understand and almost useless on its own. A Friday with an average of 14 minutes might include a quiet start at 9 minutes, a calm finish at 12 and a half hour between 8:15 and 8:45 when tickets took 28 minutes. The average hides the half hour, and the half hour is when guests complained.

Two better measures sit beside the average. The median is the middle ticket time: half of tickets were faster, half slower. It is less affected by a few extreme tickets. The 90th percentile, often written P90, is the time that 90 per cent of tickets beat: it shows how long the slowest tenth took. If the median is 13 minutes and the P90 is 26, one ticket in ten took twice as long as a typical ticket, and those tables had a different evening from everyone else.

Watch the P90 most closely. It is the number that captures the bad moments, and improving it is usually what guests notice most.

Queue time behaves in a way that surprises many kitchens. When a station has spare capacity, queue time stays close to zero. As the station approaches its limit, queue time does not rise gently; it climbs steeply, because every ticket that arrives while the cook is busy has to wait for all the tickets in front of it. A station working at nearly full stretch can go from no queue to a long one in a few minutes.

That pattern is well described by queueing theory, the study of waiting lines, and it explains why a small increase in orders at the busiest moment can produce a large increase in ticket times. It also explains why small changes help so much: taking a few tickets away from the peak, or adding a little capacity at the bottleneck, can bring queue times down sharply.

For a kitchen, the lesson is simple. Watch the busiest stations at the busiest times, and act before they reach their limit rather than after.

The single most revealing view of ticket times is by hour, or by quarter hour, through the service. Plot the median and the P90 for each interval, and the shape of the evening appears: the quiet start, the climb as the room fills, the peak when the kitchen is fullest, and the recovery.

That shape tells you when the kitchen reaches its limit. If ticket times rise sharply at the same time every Friday, the kitchen is receiving more tickets than it can cook in that interval. The answer may be pacing the dining room so fewer tables order at once, preparing more before service, or adding a cook for that hour. Our guide to pacing covers covers how to spread orders across the evening.

Compare the same day across several weeks. A single bad Friday may be an accident; four in a row are a pattern.

Some dishes take longer than others, and some take longer than they should. Measuring ticket times by dish shows which dishes slow the kitchen down most, especially at busy times. A dish that takes 12 minutes on a quiet Tuesday and 25 on a Friday is either complicated, dependent on a busy station, or short of prepared components.

Look at which dishes account for the most total time, not just the longest individual times. A dish that takes a little longer than average but sells in large numbers may cost the kitchen more time across a service than a slow dish that rarely sells. Those are the dishes where small changes in method or prep make the biggest difference.

Use the results in menu decisions. A slow dish that is also unprofitable is a strong candidate for removal; a slow dish that is popular and profitable may need its prep reorganised. Our guide to menu engineering covers how to weigh popularity against margin.

A kitchen bottleneck is the point in the kitchen that limits how fast everything else can go. It is usually one station, such as the grill, the fryer or the pass, that falls behind first when the kitchen gets busy. Every ticket that needs that station waits for it, however quickly the other stations work.

Ticket times by station reveal it. If one station's ticket times rise sharply at busy moments while the others hold steady, that station is the bottleneck. Queue times make it even clearer: a station with long queue times has more work than it can handle.

Once you have found it, you can act. Move a dish to another station, prepare more of its components before service, add a cook at peak times, or adjust the menu so fewer dishes depend on it. Then measure again, because removing one bottleneck often reveals the next.

The pass is often overlooked as a bottleneck. Even when every station keeps up, the person calling tickets, checking plates and sending them out can only handle so many tables in a quarter of an hour. When the pass falls behind, plates sit under the lights and tickets that were cooked on time leave late.

Measure the time between a ticket being marked ready at the station and leaving the pass. If that gap grows at busy times, the pass needs help: a second person, a clearer system for matching plates to tables, or runners waiting to take plates as soon as they are ready.

A clear view of every ticket at the pass, oldest first, with each station's progress visible, helps the person on the pass keep the whole kitchen in step.

All day totals are the running count of how many of each dish are currently ordered across every open ticket. If four tickets each include a sea bream, the all day total for sea bream is four. Cooks use them to batch work: grilling four fish at once rather than one at a time, or knowing that eight portions of fries are needed in the next few minutes.

Many kitchens call them out at the start of a rush: "all day, six steaks, four bream, nine fries". Some kitchen displays show them automatically. Either way, they help a station plan its work instead of reacting ticket by ticket, which can reduce cook times at the busiest moments.

All day totals also show a station's load at a glance. A station with a long all day list is about to be busy, and the pass can warn the floor before ticket times rise.

Ticket times are the kitchen's share of speed of service, but not all of it. A guest's experience of speed runs from sitting down to receiving their food, and includes the time to be greeted, to order, for the order to be sent and for the plates to be carried out. A kitchen with fast ticket times can still serve slowly if orders sit on a server's pad or plates wait at the pass for a runner.

Measure the whole journey occasionally, by following a sample of tables from seating to food arriving. Compare the kitchen's ticket times with the total, and the gap shows how much of the wait happens outside the kitchen. Our guide to order management from table to pass covers each handover.

Share the results with the floor and the kitchen together. Speed of service is a team result, and it improves fastest when both sides see the whole picture.

A target ticket time gives the kitchen something to aim for and a way to see, during service, when a ticket is running late. Targets should be realistic for the menu and the station: a cold starter might have a target of 8 minutes, a main from the grill 18, and a tasting course something else entirely.

Set targets from your own measurements rather than from a general figure. Look at the median ticket time for each dish or station on a good night, and set the target a little above it. The aim is a target that most tickets meet, so that the ones that miss it stand out.

Measure the share of tickets that meet their target, often called the on-time rate. It is easier to understand than a median or a P90, and it moves quickly when the kitchen improves.

Remakes, dishes sent back and cooked again, cost time as well as food. A remade main adds a whole cook time to the table's wait, and it takes a slot on a busy station that another ticket was waiting for. A kitchen with a high remake rate has slower ticket times than it should, even if every dish is cooked quickly.

Track remakes by dish and by reason: cooked wrongly, wrong dish, allergy issue, guest changed their mind. Many remakes come from the order rather than the kitchen, such as a missed modifier or a misheard request. Our guide to kitchen tickets that never need a shout covers how to make every ticket clear.

Measurement only matters if it leads to change. Review ticket times weekly with the head chef and the floor manager. Look at the median and P90 by hour, the stations that fell behind, the dishes that took longest and the on-time rate. Choose one or two changes for the next week, and measure their effect.

Typical changes include moving prep earlier in the day, reorganising a station's layout, moving a dish to a less busy station, adding a cook at peak, adjusting the menu, or changing how the dining room paces its seating. Each one is a test, and ticket times are the result.

Keep the record. Over a few months, it shows which changes worked, and gives the kitchen a clear picture of its own progress.

Here is an illustrative Friday service for a busy kitchen, to show how the numbers reveal a problem. The figures are examples.

Changes for the next Friday: steaks seared and rested in larger batches before 8:00, a second cook on the grill from 7:45 to 9:00, and large parties steered to 7:00 or 9:15. The following Friday, the grill's median at 8:30 is 16 minutes and the P90 is 22.

In servearound, every dish carries the prep time you set on the menu, and every ticket counts its own time in the kitchen from the moment it lands, oldest first on the board. The kitchen's live view shows how many tickets are new, preparing, ready and over their target, the oldest active ticket, and the load on each station, naming the station that is the bottleneck.

Kitchen analytics shows tickets completed, median and P90 prep time, median queue time, remake rate and on-time rate against the target you set, with filters by station, course and order type. It shows ticket time day by day and by day and hour, where the minutes go, queue depth against ticket time, which items cost the most time, and the slowest tickets with what was on them. See fulfillment and Analytics.

It depends on the dish, the station and the format. Set targets from your own kitchen's ticket times on good nights.

Ideally, yes. A kitchen display does it automatically. Without one, timing a sample on busy nights still reveals the pattern.

Because tickets are waiting longer before anyone starts them. That is queue time, and it usually points to a bottleneck or bunched orders.

Weekly, with the head chef and floor manager, and after every change you make.

They can. Time held courses from when they were fired, not when they were ordered.

Median and P90 ticket times by quarter hour, split by station. It shows when and where the kitchen falls behind.

Ticket times measure the kitchen's speed from order to pass, and they move quickly when the kitchen changes. Measure them for every ticket, split them into cook times and queue times, and read them by hour, by station and by dish. Never rely on the average alone: the median and the P90 show the moments the average hides.

Use ticket times to find kitchen bottlenecks, including the pass, and to set realistic targets. Use all day totals to help stations batch their work, connect ticket times to the whole speed of service, track remakes, and review the numbers weekly so each change can be tested against the next Friday.