Customer retention: getting guests to order again
Almost all restaurant marketing is about the guest who has never been in. The money sits with the guest who came last month and has not come back this month.
Why you keep chasing new guests
Ask an owner how the marketing is going and the answer is about new guests. Almost never about the guest who ordered twice last month and not at all this month.
That is a treadmill. You run to keep order numbers level, and the moment you stop advertising they slide. Running harder does not fix it. The guests you already have coming back more often does.
Regulars carry the profit
In most places a large share of the profit comes from a small group who order every week or every month. How large varies by restaurant, and the percentages you read come almost entirely from American research. The mechanism holds here too: you paid to win that guest once, and after that they cost you nothing.
A first order earns you very little
Off that first order come the ingredients, the wages in the kitchen, the packaging and the delivery. What is left is thin. Subtract the advert or the welcome discount that brought the guest in and there is often nothing left. It is order two, three and four where somebody starts to make you money.
Takeaway and delivery live on repeat orders
Somebody sitting at your table is more likely to add a starter and something to drink. That is where your margin is. A delivery order is usually just the main course. Take the commission Thuisbezorgd or Uber Eats charges off the top as well and you need volume. A takeaway place without regulars is quietly working for nothing.
Four emails that work while you are in the kitchen
The owners who get the most out of email are not the ones who write the most emails. They are the ones who set four up once and never look at them again. Those emails go out while the kitchen runs.
| When it goes out | What it says | |
|---|---|---|
| Feedback | One hour after the order | One question: how was it, one to five |
| Recommendation | Three to seven days later | A dish close to what they chose, with a photo |
| Repeat | After thirty days with no order | A small reason to order again tonight |
| Win-back | After sixty to ninety days of silence | Their last order, with one button to repeat it |
The feedback email is the most important and the easiest. Ask for one score, nothing more. Anyone who scores you low gets a message saying you will be in touch personally, and then you actually get in touch. Anyone who gives five stars gets asked to put them on Google. That is how you catch complaints before they become a review.
The recommendation email is the tip you would give at the counter, sent by email instead. Add a photo and one button that drops the dish straight into the basket. Every extra click costs you orders.
The discount ladder: as late and as small as possible
A discount is the easiest tool you have, which is what makes it the most dangerous. Give ten per cent every month and nobody orders without it. You have taught your guests to wait.
The way out is a ladder. You only discount once somebody goes quiet, and you start small. The longer the silence, the sharper the offer. Most guests come back on the first rung, and that rung costs least.
| Length of silence | What you send |
|---|---|
| Thirty days | Something small: free delivery, or five to ten per cent |
| Sixty days | Something bigger, ten to twenty per cent, with their last order attached |
| Ninety days or more | Your sharpest offer, once, and then you let it go |
Not every email needs a discount, and that is rather the point. A new dish, a seasonal dish that is back, a fixed evening with something special: all reasons to email without giving margin away.
Watch what you give away. Ten per cent off a 24 euro platter costs you more than free chips above a threshold you set yourself. Work from what the giveaway costs you to buy in, not from the menu price. That price includes 9 per cent VAT that was never yours.
Set the first rung up this afternoon. One email, thirty days of silence, free delivery.
Your own app is a shortcut, not a shop window
An app does not win new guests. Nobody downloads the app of a place they have never eaten at. An app is for the people who already order every week, and for them it is a shortcut: open, last order, pay. Domino’s says most of its online orders now come through its own app.
- Ordering in a few taps, with the price on the button, so nobody gets a shock at checkout.
- Favourites and a list of past orders, so a usual order is back in the basket in two taps.
- The rewards balance visible on the screen where people order, not buried in a menu.
- Photos on the dishes, because on a phone screen the photo is the menu.
Do not start on one while your ordering page is still not working. An app costs money and attention, and it only pays for itself when there is already a stream of repeat orders to speed up. When that stream is there, it is the cheapest way to hold on to it.
A rewards card only works when the balance is visible
Starbucks credits a large share of its revenue to its rewards programme. Your numbers are smaller, but the idea is the same: a guest halfway through a card picks you over the place next door.
Collect points, not a euro balance. A balance feels like money you are handing over, because that is what it is. You set what a point is worth, you can give a few away at the start, and you pick rewards that cost you very little.
American research on car wash stamp cards showed why that head start works. One card needed eight washes. The other needed ten, with two stamps already on it. The same distance from the reward, yet the second group filled the card far more often. People would rather finish what they have already begun.
Choose rewards on what they cost you, not on what they sell for. Garlic bread, a dessert, a soft drink: cheap for you, a natural partner to a main, and still worth having. A free pizza costs you the entire margin on that order.
And show the balance. A rewards card you only meet when you go looking for it does not exist as far as your guest is concerned. Show what this order does to the balance while they order.
Your ordering page decides whether there is a second time
It is easy to think of an ordering page as something for new guests. It is, but mostly it decides whether anybody comes back a second time. Good food plus an awkward order adds up to a guest who uses a platform next time. Or does not come back at all.
What somebody expects on their phone is no mystery.
- The menu as readable text, not a PDF to pinch and drag.
- A photo on the dishes that deserve one.
- Ordering without creating an account first.
- Delivery or collection, the cut-off time and the cost, clear at the top of the page.
Checkout follows the same rule: fewer screens means more orders. iDEAL and Apple Pay belong there, and an address already filled in the second time saves half a minute. Half a minute is enough to lose somebody.
Finally, put your ratings on the page. Somebody deciding between you and the place down the road looks at what the neighbourhood thinks. Your Google stars on your own page cost nothing and convince better than any sentence you write about your own food.
Three numbers that tell you whether it is working
Retention stays vague until you measure it. Three numbers are enough, and they come out of your own ordering system as long as orders carry an email address or a phone number.
| Number | How you work it out | What it tells you |
|---|---|---|
| Repeat order rate | Guests with more than one order divided by all guests, times one hundred | Whether you turn a first order into a habit |
| Customer value | Average order times orders per year times years as a guest | What you can afford to spend to win a new guest |
| Monthly retention | Last month’s guests who ordered again this month, divided by last month’s total, times one hundred | Whether your leak is getting smaller or bigger |
An example with round figures. If a thousand guests ordered last month and three hundred ordered again this month, your monthly retention is thirty per cent. On its own that number says little. Next to last month’s it says everything.
So compare yourself with yourself. The benchmarks you find online for a good repeat rate come almost entirely from American data.
Put customer value next to what a new guest costs you. Spend twenty euros on adverts per new guest while a guest is worth fifteen euros on average and you are buying revenue that loses you money. Push that value up and you can outspend the place next door, which makes it harder for them in turn.
What you can do this week
Doing all of this at once is not realistic alongside a working kitchen. Nor does it need to be. The order is: measure first, then the emails, then the rewards card and the app.
Next month you count again. If the number moves, you know where to push. If it does not, it has cost you a couple of hours rather than your margin.
Common questions
- What is a good repeat order rate for a restaurant?
- There is no reliable Dutch benchmark, and the figures you find online come almost entirely from American research on chains. Measure your own rate every month and watch whether it climbs. That is the only comparison that is actually about your restaurant.
- How often can I email my guests?
- Once or twice a month suits most places, on top of the automatic emails that follow an order. Do not only watch unsubscribes, watch how many people still open them. When that falls, you are either emailing too often or saying too little.
- Am I allowed to keep my guests’ email addresses?
- Yes, as long as you make clear at checkout that you will email them, every email carries a working unsubscribe link, and the emails are about your own food. Keep no more than you need and never sell the list on. Write down briefly why you keep what you keep: under the GDPR that is most of the work.
- Do I get the details of guests who order through Thuisbezorgd or Uber Eats?
- No. The platform keeps the customer relationship and the data. You see an order, not a guest you can email later. Somebody who orders directly from you gives their details to you. Alongside the commission, that is the real difference between the two channels.
- Do I need my own app, or is an ordering page enough?
- For most places a good ordering page is enough and an app is the step after it. An app does not win new guests: it makes ordering faster for the people who already order every week. If you do not have that group yet, build the group first.
- Do discounts work for keeping regulars?
- Discounts work to pull back somebody who has gone quiet, not to hold on to somebody who is already ordering. Send a discount every month and nobody orders without one. Use a discount as the last rung, not as a standing habit.
