How to increase restaurant revenue.

Three numbers, multiplied together. How many customers you get, what each one spends, and how often they come back. Everything below moves one of the three.

The useful thing about writing it that way is that it shows you how small the changes need to be. Take a restaurant doing 60 orders a day at €24 each. Add €1.50 to the average order and that's €90 a day, roughly €2,700 a month, near enough €32,000 over a year, from suggesting a side. Get one customer in ten to come back once more a month and you've added another few hundred covers a year without spending a euro on ads.

Owners tend to chase the first lever, new customers, because it's the loudest one. It's also the most expensive. The other two are cheaper and you already have everything you need to work on them.

Commission on the marketplace apps commonly runs 25 to 30% of the order value, and that's before card fees and any marketing spend you've agreed to inside the platform. On a €30 order that's €7.50 to €9 gone before you've paid for the chicken.

Do this sum with your own statements rather than a headline rate. Pull the last 90 days, add commission, marketing fees and card charges, then divide by the number of orders. That's your true cost per order. Most operators are surprised, and the ones who aren't surprised are usually the ones who've already stopped looking.

Then split those customers into two piles: people who found you on the app for the first time, and people who already know your name and are ordering through the app out of habit. The first pile is what you're paying for. The second pile is pure leakage, and it's the one you can fix. Put a card in every bag, a QR code on the box and a line on the receipt pointing at your own site, and give people a reason to use it. Slightly better price, a free side on a first direct order, loyalty that only counts when they order direct.

Shifting even a third of your delivery volume across is normally worth more than any other single thing on this list. There's a full comparison in direct ordering versus third-party apps.

By asking a specific question at the right moment, not by bolting a discount on everything.

"Anything else?" gets you nothing. "Do you want garlic dip with that, it's a euro fifty?" gets you a yes about a third of the time. Train the specific line for your three or four best attachments and put them on the screen at the till so nobody has to remember. On your website the same job is done by the add-on step after the main is chosen, and it never has an off day.

Drinks are where the margin lives. A can that costs you well under a euro sells for a couple of euro all night, and most kitchens under-sell them badly. Bundle sensibly too: a main, a side and a drink priced a little under the three separately is easier to say yes to than three decisions in a row.

Most menus carry a few dishes that lose money on every plate, and the owner usually doesn't know which ones.

Cost each dish properly. Weigh the portions, price the ingredients from a recent invoice rather than memory, and include the things people skip: oil, the garnish, the wrap, the box, the sauce pot, the wastage on trim. Then work in contribution margin, the cash left after food cost, not food cost percentage on its own. A dish at 35% food cost that leaves €9 on the plate beats a dish at 25% that leaves €4.

Now sort the menu into four boxes: popular and profitable, popular and thin, unpopular and profitable, unpopular and thin. Push the first box hard, at the top of the section and in the photos. Fix the second box by portion, recipe or price. Reposition the third box, sometimes just a better description does it. Cut the fourth box, because it costs you prep, stock and space on the page.

Frequency is the lever nobody works on, because it needs a customer list and most independents don't have one.

Start capturing contact details on every order that goes through your own channels, and give people a reason to hand it over. Then set two or three simple triggers and leave them running: a thank you after a first order with a small incentive to place a second, a nudge when a regular hasn't been near you in three or four weeks, an offer on a birthday, a message when a dish they order every time is back on.

Keep the loyalty scheme boring. A stamp card that lives in the system rather than a wallet, ten orders and the eleventh is on you, is understood by everyone and works. Points systems with tiers and expiry dates get ignored.

This is the sort of thing that runs on its own once it's set up, which is why marketing automation exists.

Probably, and in smaller steps than you're thinking.

Ingredients, wages, energy and insurance have all moved. If your menu prices haven't, your margin has quietly gone somewhere. Review costed dishes twice a year, not once every three, and move prices in small increments rather than one jump that everyone notices. Taking a €12.50 main to €13 is 4%. On 40 covers a night that's €20 a night, about €600 a month, and almost nobody will mention it.

A few practical rules. Don't raise everything at once, raise the dishes with the worst margin first. Leave your best-known signature dish alone for a cycle, because that's the one people remember the price of. Round to prices that read well. And check the VAT rate that applies to your service now rather than the one you remember, because hospitality rates have moved more than once and your accountant will have the current figure.

With a small, targeted, time-limited offer sent to people who already eat your food, not a permanent discount on your website.

The mistake is a standing 20% off midweek. Do that for a season and you've simply trained your Friday customers to order on a Tuesday for less money. Instead, send a modest offer at three or four in the afternoon to the segment that normally orders midweek, make it good for that evening only, and keep it small. A free side, a couple of euro off, a drink in.

There are other ways to move demand that don't touch price at all. A set lunch at a fixed number. A pre-theatre or pre-match sitting. Family trays on a Sunday. A collection-only deal that costs you no driver. Off-peak works best when it's a different product rather than the same product cheaper.

Look at the hours and the assets you're already paying for. The kitchen is warm, the staff are rostered, the rent is the same on a wet Monday.

Office and site catering is the obvious one for a takeaway: trays, a fixed price per head, ordered the day before, collected or dropped in one run. Gift cards turn into cash now for food served later, and a share of them are never redeemed. Pre-orders for a big match, a bank holiday or a local event let you buy accurately instead of guessing. If you make one thing well and it travels, a jar or a chilled portion at the counter is found money on the way out the door.

None of these need new equipment. They need a price, a cut-off time and somebody to answer the phone.

Some of your missing revenue isn't missing, it's spilling. It costs nothing to plug and most owners have never measured it.

Count the calls you miss between six and nine. Every unanswered phone at that hour is an order that went to the restaurant across the road. Then open your own checkout on a mid-range phone on mobile data and order something. Time it. If it takes more than a minute or asks a hungry person to create an account, you're losing baskets you already paid to attract. Check your opening hours on Google as well, because a wrong bank holiday listing quietly closes you on your busiest day.

Pick a handful of numbers and look at them every week, on the same morning, for long enough to see a trend.

Revenue by channel matters more than total revenue, because the mix decides what you keep. Average order value tells you whether the upselling is real. Repeat rate over 90 days tells you whether the win-back messages are doing anything. Contribution margin per dish tells you whether the menu changes worked. And revenue per opening hour will tell you, honestly, whether that Monday lunch service is worth opening the door for.

Half an hour on a quiet morning beats a grand quarterly review that never happens. There's more on which numbers matter in the restaurant analytics guide.