Raising Menu Prices Without Losing Regulars: How Many Customers You Can Afford to Lose

Your costs have moved and the menu hasn't changed in a year, and every time you think about raising menu prices the same question comes back: how many customers will I lose? That answer isn't a feeling. It's a number you can work out before you change a single price. In this guide we go through the figures you line up first, the simple math that tells you how big a drop in orders leaves you exactly even, and a three-batch rollout that measures each round before you open the next one.

Let's be straight from the start: there's no method that stops people noticing. There's a method that keeps the people who notice buying anyway, and that lets you tell a dip caused by your price apart from a dip caused by the season, so you don't reverse a decision that was right because one week came in bad.

Before you touch a single price: the numbers that decide it

Raising prices without these figures is pricing by luck, and then explaining the result by luck too. Five numbers per item you're thinking of touching, all of which you already have if your sales and costs are recorded:

• Recipe cost per item: what the dish actually costs you by ingredient and weight, not what you reckon in your head. When cost is built at recipe level, a supplier price change flows into it on its own. How to build that is covered in recipe costing and food cost percentage.
• Contribution margin: the item's price with the tax stripped out, minus its recipe cost. That's the number paying rent, wages and power, not the food cost percentage on its own.
• Monthly units sold per item: an item doing 300 units is a completely different decision from one doing 20, even if the percentage you're adding is identical.
• Average check and invoice count: both together, so that later you can tell a drop in customer numbers apart from a drop in what each order is worth.
• Repeat-customer share: someone who comes in weekly holds your prices in their head far better than a passer-by, and they're the ones who give you the first honest signal after a change.

And here's the detail plenty of people get wrong: consumer prices here are shown with VAT included, so the menu price is not your revenue. If a dish is 46 on the menu, the net that reaches you is 40, and the rest is tax you're collecting on someone else's behalf. Any margin math done on 46 instead of 40 shows you profit that doesn't exist, and pushes you into raising either too little or too much. Rules on price display and invoicing change, so check the Zakat, Tax and Customs Authority for your own case before you lean on any detail here.

How many customers can you afford to lose? The math that settles it

A worked example to redo with your own figures: a dish listed at 46 including tax, net 40, recipe cost 14. Every unit gives you 26 of contribution. You sell 300 a month, so that's 7,800 in monthly contribution. You decide to move it to 50 on the menu, which puts the net around 43.5 and the contribution at 29.5. To land on the same 7,800 you now only need 265 units. In other words you can afford to shed 35 units, about 12% of that item's sales, and still be exactly even. Anything smaller than that and you're ahead. That's the number you run on instead of guessing.

Read the table backwards to get it right: the higher your margin, the less room you have, because every unit you lose takes a bigger contribution with it. And the more you raise, the wider the gap you're allowed to lose. Which means a small bump on a high-margin item is riskier than it looks, and a considered bump on a thin-margin item is far gentler. These figures are illustrative. Put your own in and redo it item by item, because an item at 45% margin has nothing in common with one at 75% when it comes to this decision.

The three-batch rollout

One sweep across the whole menu muddles everything: sales dip and you have no idea which item did it, so you either roll the whole thing back or dig in on the whole thing. The alternative is three spaced batches, each touching one group of the menu and measured on its own before you open the next. This is risk management, not a promise nobody will grumble. It just means you'll know exactly where the grumbling is, and you can reverse a part instead of the lot.

Levers to try before you touch the price

Sometimes the problem isn't the price at all, it's the cost sitting under it or the leak sitting beside it. Before you open the menu and start raising, you've got four levers that get you part of the same effect without the customer feeling a thing. Work them in order and see how much of the gap closes before you even need an increase:

• Revisit the most expensive ingredient in the recipe: open the recipe cost and see which ingredient takes the biggest share of the dish. Negotiating it or changing your purchase size can move your margin more than a 5% price bump would, and with no reaction from anybody.
• Fix your portion weights: an item costed at 14 on paper and plated at 17 in practice loses you three on every unit, which is a full price increase you're charging yourself daily. Written weights and a scale on the line claw that back with no price change.
• Price add-ons and sizes on their own: instead of moving the base item, let the larger size and the extras carry a price that reflects what they actually cost. The average check rises while the headline price on your best-known items stays put.
• Close the leaks: waste, unrecorded staff meals, and open-ended discounting at the till. Every unit that leaves without an invoice does precisely what the sales drop you're afraid of would do, so put a logged reason and a permission behind every discount.

Segment the menu before you touch a price

Items don't share the same price sensitivity. Some, customers hold the price of because they buy them every visit. Others they buy once a month and couldn't tell you what they cost. To find the price-anchored items in your business, rather than by some general rule, pull the ranking of best sellers by unit count instead of by revenue, and see which of those keep showing up on repeat customers' invoices. There's also a quick tell you pick up standing at the till: anything a customer orders without opening the menu is a strong candidate for the price-anchored list, and you confirm it off the frequency ranking rather than off your impression. That ranking by unit count comes out of the sales report in whatever till system you run, or off four weeks of receipts totted into a sheet if you haven't got the report.

• Price-anchored items: the ones appearing most often on repeat customers' bills. Leave them for the last batch and move them the least, or leave them alone and make it up elsewhere.
• Thin-margin items: contribution below your menu average, and selling well. That's your first batch, because a change here shifts the result faster than anywhere else.
• Add-ons and sizes: extra cheese, the large size, the extra sauce. Pricing add-ons lifts the average check without touching the headline price of the main item.
• Rare items: anything under a threshold you set for your own size, say ten units a month. Raising them changes nothing in your revenue, and cutting them frees your kitchen more than a price change ever will.

And watch for a side effect that tends to slip past: when you raise something in the middle of a section, some customers don't walk away, they drop to the cheaper item sitting next to it in that same section. The result is a flat invoice count with a lower average check, while you tell yourself you lost nobody. So don't measure the item you raised on its own, measure the whole section: total units and total contribution for the section, before and after. Flat units with lower contribution means your increase pushed customers towards a choice that pays you less, and that loss never shows up in a single-item report.

Once the menu is segmented, work through this order and write each step down with its date, because the whole value of the plan sits in the measuring between the rounds:

1. 1
   Lock a baseline before you change anything

   Record per item: units, contribution, average check and invoice count, over the last four full weeks. Without a written baseline, any comparison later is memory rather than measurement, and memory always leans towards the decision you already made.

2. 2
   Price batch one: the thin margins

   Move three to five items whose contribution sits below your menu average, and keep each item's increase inside what you can justify to yourself off the table above. Do not touch the price-anchored items in this round at all.

3. 3
   Change the price on every channel the same day

   Till, printed menu, online store, table QR and the delivery app menus. A price that differs between channels reaches the customer before it reaches you, and turns an ordinary increase into a complaint you didn't need.

4. 4
   Wait four weeks and compare like days

   Thursday against Thursday, Tuesday against Tuesday, not month against month. Drop any week with a season, a holiday or an event in it, so you don't charge the season's effect to your price and walk out with the wrong call.

5. 5
   Read three signals, not one

   Units of the item you raised, average check and repeat-customer share. Units falling while the average check holds means customers switched to something else on your menu. Units falling alongside invoice count means you lost visits. The difference flips your decision completely.

6. 6
   Decide: continue, hold, or reverse

   If the drop came in under the ceiling your math gave you, open batch two six to eight weeks later. If one item blew past the ceiling, put that single item back and leave the rest, and log the reason and the date so you don't repeat the same experiment next year having forgotten how it went.

The timing itself matters as well. Don't raise in a week carrying a season or an event, because the natural lift papers over the price effect and hands you a false read you'll then build the next batch on. Don't raise the week after a service problem or a run of complaints either, because customers will tie the two together even when they're unrelated. The cleanest slot is an ordinary week, neither the start nor the end of the month. And if your menu is printed, let the print cycle set the date of batch one so you're not paying for two runs, then take the remaining two batches across screens and digital channels where changing a price costs you nothing.

Delivery app prices are a separate story

An order coming through a delivery app has commission taken out of it, typically somewhere between 15% and 30% depending on the app and your agreement, which means the same dish at the same price leaves you a far thinner margin than a dine-in order. So before you raise everything to make up for one channel, split the math: work out the item's contribution in the dining room and its contribution on the app separately. Sometimes a different price on the app menu is cleaner than a blanket increase, bearing in mind that some agreements carry a price-parity clause, so read your contract first. And there's a third route that doesn't touch prices at all: shifting some of those orders onto your own channels, which take no commission per order, and the per-order commission math is worked through in full in managing delivery app commissions.

A worked example on the same dish as above: priced at 50, net 43.5, recipe cost 14, so dine-in contribution is 29.5. If commission on the order runs at 20%, that takes 10 out, and the same dish now contributes 19.5 through the app instead of 29.5. A gap of 10 on every unit, and that single figure is what tells you whether you need a different price for the channel or you need to move some of the orders. How commission is calculated varies with your agreement, and some of it is charged on the full order value, so confirm the basis in your contract before you build any pricing decision on it.

What the person on the till actually says

A well-planned increase gets wrecked when it reaches the customer in a flustered sentence. Write one line for the whole team, honest and short, along the lines of: our ingredient costs moved, we updated some items, the rest are unchanged. No long apology, no supply-chain lecture, because a customer who asks wants a confident answer rather than a justification. And the day before the change, walk the team through it so they hear the new number from you, because hesitation at the till reads to a customer as an increase with no reason behind it.

• Don't announce the increase: update the menu and move on. An announcement draws attention to an item that would have passed unremarked.
• One price everywhere: print, screen, app and store. A mismatch gets read as bad faith even when it's an oversight.
• Don't raise the price and shrink the portion together: two changes on the same dish in the same week leave you unable to tell which one drove the reaction, and they break the measurement the whole plan rests on.
• Land on clean numbers deliberately: rounding down to a comfortable figure eats part of your increase, so run the math on the rounded price, not the one before rounding.

Once the three batches are done: what good looks like

After three batches with measurement in between, you should have a menu whose prices are built on recipe cost and margin rather than on the last time the place next door moved. More valuable than the menu itself is the record you now hold: for every item, what it was, what it became, when it changed, and what happened to its sales in the four weeks after. That record is what makes the next increase quicker and calmer, because you're not starting from zero or from an open argument. Keeping it is far easier when recipe cost and per-item sales come out of one system rather than scattered files: all the features in detail. And if you'd rather rebuild pricing from the ground up than patch today's numbers, start from how to price menu items.

• Contribution per item at the new cost: supplier prices move, so the margin you calculated today wants revisiting every three months.
• Average check: if it fell after the increase, customers are buying less rather than disappearing, and the fix for that is a completely different one.
• Repeat-customer share: the slowest signal and the most honest. A sustained slide after an increase is worth a pause before the next batch.
• Count of items selling under ten units a month: the bigger that number gets, the heavier your menu sits on the kitchen for nothing in return.

And set yourself a written reversal rule before you raise anything: any item whose units fall past the ceiling your math gave you goes back to its old price within two weeks, and stops there. A decision written before the increase protects you from an emotional one after the first bad week, and a bad week is always on the cards for reasons that have nothing to do with price: weather, a closed road, a holiday, or a competitor opening next door.