More online sales, channel by channel

Sales come from three levers, not one

Ask an owner how they plan to grow this year and the answer is almost always the same: more guests. That is the most expensive of the three levers you have, and usually the last one you should touch.

The other two are already inside your restaurant. A guest who comes back every month instead of once a quarter is worth three times as much, and you spend nothing on advertising to get there. An order with a drink attached earns you more margin than the dish on its own.

• More people finding you. Costs money or time, and works slowly.
• More people coming back. Costs almost nothing, as long as you know who they are.
• More value per order. Costs nothing and works from tomorrow.

Underneath all three sits a fourth question that rarely gets asked: which channel does the order arrive through. Two twenty-five euro orders are not worth the same if one comes through a platform and the other through your own site. On the first you pay commission and you do not know who ordered.

The rest of this piece walks those channels: where your sales come from, and what there is to win in each.

Be findable at the moment someone is hungry

Most orders do not begin with you. They begin with someone hungry picking up a phone and typing what they want to eat and where they are. Whatever sits at the top gets the order.

That makes search the highest intent channel there is. Someone typing kebab delivery Utrecht does not need convincing that they are hungry. They want a menu, a delivery time and a button. If they cannot find those quickly, they will find them two lines further down.

Where to start

• Your name, address, phone number and opening hours exactly the same everywhere: your site, your Google Business Profile, every directory.
• Your guests’ words, not the ones on your shopfront. Pizza takeaway Rotterdam Noord is not pretty, but it is what people type.
• Your menu as ordinary text, not a PDF. PDFs read badly on a phone and a search engine gets little out of them.
• Describe your photos in the code: salmon with lemon butter sauce, not IMG_4471.
• Ask for Google reviews and reply to everything that comes in, including the unpleasant ones.

Reviews do two jobs at once. They help you climb the map and they convince the guest who has already found you. The same holds on TripAdvisor and Iens, but Google is where most people look first.

Do one thing on Monday: search for yourself on your phone in a private window and look at what a stranger sees. Inside a minute you will know what is missing.

Your website is your busiest table

Your site is open all day, including the day the restaurant is closed, and more people walk past it than could ever sit inside. It still gets less attention than anything else.

A visitor wants three things answered within seconds: can I order here, what can I get, and when does it arrive. Everything standing in between costs you orders. A full-screen mood shot, background music, a newsletter pop-up before anyone knows what you sell.

What has to be on it

• Your menu as readable text, by category, with prices and photos.
• An order button visible on every screen without scrolling.
• Opening hours, delivery area and delivery time, without anyone having to ring.
• Your address with a map link, and a phone number somebody answers.
• A short page about who you are: the wood oven, your mother’s sauce, the year you opened.

That last one looks the least useful and is most often the difference. Two restaurants at the same price with the same delivery time do not beat each other on price. They win on the reason someone would rather order from you, and that reason has to be readable somewhere.

On an ordering site, photos are not decoration. They are the reason someone picks a dish they have never had, and that is more often the dearer one. Start with the ten you sell most and the three you make the most margin on.

A slow site is a closed door

Speed is the dullest subject in this piece and the most expensive one to leave alone. A hungry person staring at a blank screen does not wait. They go back to the tab next door, usually a platform they are already logged into.

You can measure it without understanding any of it. Put your address into Google’s PageSpeed Insights and you get a report on what is slowing your site down.

• Photos straight off a phone, ten times larger than needed.
• Third-party scripts: a chat window, two pop-ups, three sets of analytics.
• A slow server, usually the cheapest hosting package.
• Plugins installed for something once and left there.

The first is nearly always the biggest and the only one you can solve without help. If you do one thing, shrink your photos.

The channel decides what is left of an order

Thuisbezorgd, Uber Eats and Deliveroo put you in front of people who would never have found you on their own. That is genuinely worth something. You just pay for that visibility per order, and you keep paying for the same guest who has been ordering every month for two years.

The point is not that you have to leave the platforms. The point is that you have two channels with different arithmetic, so you use them for different jobs.

So use the platform as a shop window and your own site as the till. New guests are welcome there. Just make sure that by the second order they know direct ordering exists, and that it is every bit as easy.

Work it out once a month in euros. Not the percentage, the amount that was withheld. That is what your own channel is allowed to cost per month before it starts saving you money.

Without data, every order is a first order

From a platform order you learn what was ordered and almost nothing else. No name you are allowed to use, no email address, no phone number. So you can never reach that guest on your own. If they come back, it is because the platform reminded them, and it is not reminding them only about you.

Through your own site you do have those details. Over time that is worth more than the difference in commission.

• Ask for permission to email or text at checkout. One tick box, not a form.
• Keep a record of who ordered what. That is the start of any offer that makes sense.
• Send little: a new menu, a holiday, a week you are closed.
• Stay within the rules: ask for consent, make unsubscribing easy, sell nothing on.

A loyalty programme or a targeted offer only works if you have both of those things: a list and a history. Without them, every promotion is an advert to strangers.

Get more out of the orders you already have

The cheapest lift in sales is on the screen where the guest is already checking out. They have decided they are eating with you tonight. The only open question is how much.

Put the suggestion where it belongs. Offering a soft drink at checkout works, putting it at the top of the menu does not. Extras belong with the dish itself, not in a category at the bottom that nobody opens.

Choose what you offer on margin, not on price. Hospitality VAT here is 9% on food and soft drinks and 21% on alcohol, so what you keep from a beer is less than the price suggests. Work out what a combination earns you once, before you fix it in place.

• A combination that feels cheaper than the separate parts, and is not cheaper on your margin.
• A second size of the same dish, so the guest trades up on their own.
• An extra you make yourself that costs you almost nothing.

How to set that up in detail belongs in the piece on average order value. This is the fastest lever to turn, because it needs no new guest at all.

Let your neighbourhood do the work

Reach does not have to be bought. Every town has a network sitting there that costs nothing, and most restaurants leave it alone because it is more effort than switching on an advert.

• The office around the corner. A standing lunch arrangement is predictable revenue in your quietest hours.
• The brewery or the cheese shop down the road. Doing something together costs nothing and gives both of you reach.
• The local paper and the city blog. They have space to fill every week, and a new menu is a story to them.
• Food bloggers from your own city. Cheaper than advertising and more readily believed.

Start small. One partner you already talk to, and only expand once you can see something coming of it.

Gift cards belong in the same list. Someone buys one for someone else, that person walks in without you going looking for them, and whatever the bill runs over the value of the card they pay themselves.

Pick three things and measure them for four weeks

Doing everything above at once is not a plan. It is a way of not knowing, four weeks from now, which part worked.

Pick three: one for how findable you are, one for your site and one for your channel. Switch them on, leave them alone for a month, and only then look. Four weeks is short enough to sustain and long enough to show something.

The last two matter most and are tracked least. Commission as a percentage feels like a rate you cannot do anything about. Commission in euros is an amount, and it can sit next to the monthly price of your own ordering system. Once those two numbers are side by side, the rest is arithmetic.