Why is my restaurant struggling?

This is the most common reason a busy-looking restaurant still loses money. When 25 to 30% of every delivery order leaves in commission, and card and service charges and funded promotions sit on top of that, a kitchen that looks full all weekend can still miss the rent.

Get the real number rather than the headline one. Pull ninety days of platform statements, add every deduction, and divide by the orders you actually received. Then compare that with your gross profit per order. It's common to find the busiest platform is the one you make least on, and occasionally to find one where you're working for nothing.

Then split those customers in two. People who genuinely discovered you in the app are what the commission buys. People who already knew your name and are ordering through the app out of habit are pure leakage, and that's the group you can move. Put a card with a QR code in every bag, an offer worth switching for on the first direct order, and loyalty that only counts on your own channel. There's a fuller version in direct ordering versus third-party apps.

A first-time customer who never returns is one of the most expensive things in this trade, and most restaurants have no idea how many they have.

If you aren't capturing a name and a contact detail, every marketing euro goes into finding brand new people, which is the hardest and dearest kind of growth. Worse, when your orders run through marketplaces, the platform holds the relationship, not you. You can't tell those customers you've added a dish, moved premises or changed your hours, and if the app changes how it ranks restaurants tomorrow, your revenue changes with it.

Start narrow. Capture an email or number on every order through your own channels and ask at the counter. Then set two or three triggers and leave them: a thank you after a first order with a reason to place a second, a nudge when a regular has been quiet for three or four weeks, something on a birthday. Even a rough version of this usually lifts repeat orders inside a month, and unlike advertising, it keeps working after you stop paying attention to it.

If you don't appear in the map results when somebody nearby searches your cuisine, you're invisible at the exact moment they decide.

Claim your Google Business Profile and fill it in properly: categories, exact hours including bank holidays, delivery and collection, the menu, and an order link that goes to your own site. Add real photos of your food and your room, taken on a phone near a window. Reply to every review in a normal voice, including the bad ones, and never argue in public. Ask for reviews when somebody tells you the meal was good on the way out, not in a mailshot a fortnight later.

Check the boring things too, because they cause more damage than they should. Wrong opening hours on a bank holiday closes you on your busiest day. An old phone number sends orders nowhere. A website that isn't linked, or that points at a marketplace, hands the customer straight back to a platform charging you a quarter of the order.

The uncomfortable one, and worth going through honestly before you spend money on anything else. Sometimes the restaurant is quiet because the last few visits weren't good enough.

Read your last fifty reviews in one sitting, in order, and count the themes rather than reacting to individual ones. Cold food, long waits, small portions, wrong orders and rude service each point at a different fix. Then order your own food to your own house at half seven on a Friday, unannounced, and eat it the way a customer does. Delivery is where consistency dies: a dish that's excellent at the pass can arrive twenty minutes later with steam-softened batter and no heat left in it.

Check consistency across shifts as well. If the Saturday chef's version of your best-selling dish differs from the Tuesday one, that's a recipe and training problem, not a marketing problem, and no amount of promotion will paper over it. Fixing the product first is slower and less fun than launching a campaign, but a campaign that brings people back to the same experience only speeds up the decline.

A slow, awkward or missing ordering page is the easiest way to lose orders you've already paid to attract.

Test it properly. Not on the office wifi on a laptop, but on a mid-range phone on mobile data at seven on a Friday. If it takes more than a few seconds to load, insists on an account before anybody can order, hides the delivery time until the last step, or doesn't offer Apple Pay and Google Pay, people close the tab and open a delivery app instead. They won't tell you. They'll just never appear in your numbers.

Fix the basics in this order: fast mobile load, guest checkout with details remembered for returning customers, one-tap reorder of a previous order, delivery time and full total shown up front, and prices that match what people see in the restaurant. If you can track it, look at how many baskets get started against how many orders get placed. That gap is money you've already spent to attract.

Most owners find out something went wrong days later, sitting down with the paperwork on a Sunday. By then the slow Thursday is gone, the spike in voids has become a habit, and the dish that ran out at eight has already cost you a service.

You need a short daily look and a proper weekly one. Daily is two minutes: takings by channel, order count, average order value, anything obviously odd. Weekly is half an hour on a quiet morning, comparing this Tuesday with the last four Tuesdays rather than this month with last month, because months have different numbers of Fridays and one of them had a bank holiday in it.

Compare like with like and the picture stops lying to you.

Ingredients, wages, energy and insurance have all moved. If your menu hasn't, the margin has gone somewhere, one euro at a time, and the takings look fine right up until the accountant calls.

Re-cost your dishes at least twice a year. Weigh the portions as they actually leave the pass, price the ingredients from a recent invoice, and include the packaging, the sauce pot and the trim waste. Work in cash margin per plate rather than food cost percentage on its own. Then move prices in small steps rather than one jump everybody notices: taking a €12.50 main to €13 is 4%, and on 40 covers a night that's about €600 a month that nobody will mention.

Two more things worth doing at the same time. Raise the worst-margin dishes first and leave your best-known signature price alone for a cycle, since that's the one people remember. And check the VAT rate applying to your service now rather than the one you remember, because hospitality rates have changed more than once and your accountant will have the current figure.

Gut feel is a fine starting point and a poor long-term strategy. If you can't say which day, hour and channel make you money, and which quietly lose it, every decision is half blind.

You don't need a data team, you need a weekly habit. Half an hour on a quiet morning: sales by day, sales by hour, top and bottom items, average order value by channel, and the share of orders coming through your own site rather than the apps. Compare with the same period. Write down one question about anything that moved, then go and answer it by standing on the floor at that hour.

Then do one thing about it before the next review. One, not five. Most turnarounds start with an honest look at the numbers followed by a single change that actually gets finished.