How many restaurants actually make it, according to the data

The number everybody repeats

You hear it at the trade fair, at the bank and at the birthday table. Ninety per cent do not survive year one. Sometimes it is five years, sometimes ninety-five per cent. The number moves, the tone does not.

It never arrives with a source. As far as anyone can trace it, it surfaced in an American television advert, and was repeated until it sounded like research.

That is not harmless. It is a hard trade, and everyone in it knows that. But start from the assumption that it will go wrong anyway and you behave accordingly: you do not invest in your own ordering channel and you accept commission that eats your margin.

What the data actually says

The best evidenced figure is American. Researchers used data from the US Bureau of Labor Statistics to work out how long restaurants there stay open. Not ninety per cent in the first year, but around seventeen.

So in that measurement restaurants did slightly better than the rest of the service sector. Not because the trade is easy, but because the myth sits so far from the measurement.

Be honest about this: it covers American businesses in an American market, with different rents, wages and delivery platforms. No Dutch figure has been measured the same way.

A restaurant carries ordinary business risk. Below are the five reasons places actually close, and not the obvious ones.

1. Your menu does not fit the neighbourhood.
2. Nobody can say in one sentence what your place is known for.
3. Your margin leaks before you see it.
4. Guests come once and never again.
5. Your online shopfront turns people away before they are through the door.

Reason 1: your menu does not fit the neighbourhood

Menu-market fit is an ugly phrase for something simple: match what you sell to what your area wants and is not being served well. A good place closes on the wrong street.

First look for demand nobody serves

1. Open Google Keyword Planner, free with a Google Ads account, with your city or neighbourhood as the location.
2. Type in your concept and the dishes you want to be known for: poke bowl, Turkish pizza, ham roll.
3. Count how many places nearby sell exactly that and read their reviews.

People searching Google for food usually have no regular place yet. High volume and few places doing it well is a gap. Many places and little volume is a fight you can let go.

Then read your competitors' reviews

Order from the four places closest to your plan, read their reviews on Google, TripAdvisor and Iens, and note what you find.

• Product: is quality consistent, or is it always the same dish that disappoints?
• Place: how easy is it to get there, and can you order direct or only through Thuisbezorgd?
• Price: does the price match what lands on the plate?
• Promotion: do guests mention how they found the place?

An afternoon of reading gives you more than a market study you commission.

Reason 2: nobody can say in one sentence what you are

Your brand is not your logo. It is what people think when your name comes up, and what they tell a friend when they recommend you.

The strongest places are sharp about something. The weakest sit in the middle: decent food, decent price, decent atmosphere, no reason to choose them.

• You are not known for anything. Anyone who cannot recommend you in one sentence will not.
• Your menu gets out of hand. Twenty more dishes means more stock, more waste and fewer dishes you are genuinely good at.

Choose your guest first, the look second

Do not start with colours, but with the regular who spends the most and recommends you for the right reasons. Write down who that is and why.

Your brand lives in what you sell, what you say and how you serve. When those three tell the same story, a guest understands your place at once.

Reason 3: your margin leaks before you see it

Whoever can spend the most to win a guest wins over time. That is why margin matters: you have to make money on a dish after service, packaging and marketing come off it.

Cutting quality almost never works. Guests forgive a great deal as long as the food is right.

First know which margin you have

Work with revenue excluding VAT. Food carries nine per cent, most drinks twenty-one. Skip that step and your margin looks better than it is.

Then take the commission out

The most direct lever is more orders arriving direct. Same pizza, same box, same rider, only without the commission. That difference repeats every month.

Then raise the size of the order. A common rule of thumb is three times your food cost, but on an expensive main that gets out of hand. Turn it round: price the main at the top of what your area pays and add small things that are nearly all margin.

Garlic bread with the pasta, an extra sauce, a soft drink. Put them in your online menu under the dish they belong to and train your staff to mention them. That is profit on an order you already had.

Reason 4: guests come once

Regulars carry a large share of the profit. How large varies by business and by source, but the direction never changes: a guest you already have costs you no advertising money.

And people usually do not stay away because they were unhappy. They were busy and they forgot you.

First capture who ordered

Ask for an email address at checkout, and a phone number if you use one. Those are channels you own, with no algorithm deciding whether your message lands.

Under the GDPR you ask permission before sending marketing, and every message carries an unsubscribe link. Platform orders usually do not give you those details at all, which makes commission dearer than the percentage printed on it.

Then automate two messages

• A few days after a first order: thank you, and here is what people usually add.
• When a regular drifts off: three orders and then six weeks of nothing, so one message with what they chose.

Then make coming back easy. A regular wants to order the usual in three taps, and a saved order history does more for that than a discount. Use points rather than a lump sum, and reward with something that carries a good margin.

Reason 5: your online shopfront turns people away

For takeaway and delivery your website is your front door. Guests look online first and decide whether they order from you or from the place two streets down.

You have a few seconds, and they answer two questions: do I understand this, and can I order easily.

The homepage

• One sentence at the top saying what you are and where. ‘Wood-fired Neapolitan pizza in De Pijp’ says enough.
• One large, sharp photo of your best known dish. An overhead shot of a plate sells better than your façade.
• An order button that still looks like a button on a phone.

The ordering flow

• Payment by iDEAL, plus Apple Pay or Google Pay for people who default to it. Every extra step costs orders.
• A photo on every dish, not only on the homepage.
• Your Google rating next to your most popular dishes. A new guest wants to know whether the neighbourhood likes it.

Open your own site on your phone, out in the street, on mobile data. Order something and pay for it. Whatever goes wrong there goes wrong for your guests too.

What you can do this week

Five reasons are four too many to start with on Monday. Pick one.

None of the five saves your business on its own. Together they decide which column you are in five years from now.

And when somebody tells you that ninety per cent of restaurants do not make it, you now know what to ask. Where is that written down?