What is order throttling? A Zayos guide to pacing a kitchen that is running hot

Published July 2026 · Updated July 26, 2026

Order throttling is a kitchen limiting how many online orders it accepts in a set window, usually per fifteen minutes. Instead of taking every ticket the moment it arrives, the system holds or delays new orders so the line can cook what it already has.

Also called order pacing or order capping. Every delivery platform has some version of it, and most kitchens have never opened the setting.

Why should an owner care about pacing at all?

Because an over promised order costs more than a refused one. A kitchen that accepts everything on a Friday pays for the food, pays the labor to cook it, pays the packaging, and then eats the refund, the one star rating and the customer who never comes back. A kitchen that paces the same rush cooks slightly fewer orders and keeps every one of them.

A late order is more expensive than a declined one. You paid for the food, the labor and the packaging, and then you refund the ticket and lose the customer as well.

There is a second cost nobody puts on a P and L. Ratings are sticky. A bad week of quoted times sits in your average for months, and on a marketplace your average rating is a ranking input, so a rough Friday keeps costing you placement long after the rush is over.

What does it look like when a kitchen does not throttle?

One Friday, seventy minutes, no pacing anywhere. The times are illustrative. The sequence is what every operator recognizes.

  1. 6:40 pm

    Six tickets on the board

    The line is clearing them in about twelve minutes each. The apps are quoting diners twenty five minutes and that quote is honest.

  2. 7:05 pm

    Fourteen tickets

    Orders are arriving faster than the line clears them. The quoted time has not moved, because most platforms quote from a fixed prep time you set once and forgot.

  3. 7:20 pm

    Twenty two tickets

    Real time to the door is now around forty five minutes. Every new diner is still being promised twenty five. The gap between the promise and the kitchen is the entire problem.

  4. 7:35 pm

    The calls start

    Drivers are waiting in the doorway for food that is eleven tickets away. Two orders cancel. The food for one of them is already cooked.

  5. 7:50 pm

    The night is decided

    The kitchen is cooking food nobody is waiting for anymore, and the ratings for the evening are already written. Nothing that happens at 9pm fixes it.

Nothing in that hour was a cooking failure. The line was fast the whole time. What broke was the distance between what the diner was promised and what the kitchen could actually deliver, and pacing is the only tool that closes it while service is still running.

What are the four ways to throttle orders?

In order of how much you should reach for them. The first two are daily tools. The last one is an emergency brake.

01 Prep time buffer

Raise the quoted prep time so the promise matches the kitchen. Blunt, instant, and the safest of the four, because it slows the arrival rate by making the quote honest rather than by rejecting anyone. Most kitchens should be adjusting this twice a day, not twice a year.

Costs you the diners who compare quoted times and pick the faster kitchen.

02 Order caps per window

A hard limit of, say, twelve orders per fifteen minutes. Order thirteen is held and released into the next window. Precise, and it requires an honest capacity number rather than an optimistic one.

Costs you nothing if the cap is right and real revenue if the cap is set too low out of caution.

03 Item level pacing

Cap the specific items that bottleneck one station instead of capping the whole kitchen. If the grill is the constraint, limit mixed grill platters to four per fifteen minutes and let the wraps flow. This is the lever most kitchens never touch and the one that usually helps most.

Costs some ticket average on your highest priced items, which is why you pace them rather than remove them.

04 Channel pause

Stop new orders on one channel for ten or twenty minutes. This is the emergency brake, not a strategy. If you reach for it more than once a service, the real answer is one of the three levers above or a prep change.

Costs marketplace placement, because platforms rank kitchens that keep going ahead of kitchens that keep stopping.

When does throttling cost you money?

Throttling is a fix for a spike, not a substitute for prep. If you are capping orders every single service, the cap is hiding a staffing or prep problem that is quietly costing you more than the orders you are turning away.

Three specific ways it goes wrong. First, capping by ticket count treats a nine dollar side the same as a twelve item family order, so pace by items or by station minutes wherever the platform lets you. Second, leaving a cap on after the rush ends means you spend your quiet hour declining the orders you actually wanted. Third, and the most expensive, throttling the channel you own before the channel that charges you 25 to 35 percent blended. Shed the expensive revenue first.

If you have to shed volume, shed the most expensive channel first. Your own site is the cheapest order you will take all night.

Where Zayos fits

Zayos direct orders land on the same kitchen tablet the restaurant already uses for its marketplace tickets, so pacing decisions get made in one place instead of across three apps with three different settings. Plans are $499, $599 and $699 per location per month, month to month, no setup fee, and the restaurant keeps 100 percent of food revenue and tips. The diner pays a service fee at checkout: $0.99 pickup, $2.99 delivery, $0 dine in, 10 percent on catering.

How the channels consolidate is covered on the Otter integration page and in how it works. More kitchen and platform terms are defined in the glossary.

See what a busy Friday actually pays you

The free AI report runs your own order volume and fee stack in about sixty seconds. No card.

Get my free AI report → Start now

Throttling questions from the pass

How do I work out the right throttle number for my kitchen?
Count what the line clears in its best fifteen minutes, not its average fifteen minutes, then set the cap one or two tickets under that. A line that has genuinely cleared fourteen tickets in a good quarter hour should be capped near twelve, not near eight. Recount it every time the menu or the staffing changes, because the number is a property of your current line and not a permanent setting.
Does throttling hurt my ranking on the delivery apps?
Pausing a channel does, because platforms rank availability and a kitchen that keeps going gets shown ahead of one that keeps stopping. Honest prep times and item level pacing generally do not, because the store stays open and the orders keep flowing. Rank the levers that way: buffer and item pacing first, hard caps second, channel pause last.
Should I throttle orders from my own website too?
Only after the marketplaces are throttled first. Direct orders are the cheapest revenue you have, since there is no commission on them, so they should be the last thing you shed when the line is under pressure. Turning off your own site while leaving a 25 to 35 percent blended channel wide open is backwards, and it happens more often than you would think.
Is order throttling the same as pausing my store?
No. Pausing stops everything and makes you look closed. Throttling keeps you open and moves the overflow a few minutes later. A paused store loses the order permanently. A throttled order arrives, gets cooked, and simply lands in the next window.
How do I know my kitchen needs throttling at all?
Compare the quoted time on your busiest night with the real time from ticket in to bag out. If the gap is under five minutes you have a healthy line. If the quote says twenty five and the reality is forty five, you do not have a speed problem, you have a pacing problem, and the diners in that gap are the ones writing your reviews.

Free report, sixty seconds, no card: grader.zay-os.com.