How do I win back customers who stopped coming?

How to win back customers who have stopped coming is one of the questions we hear most across the 10,000+ restaurants we work with around the world, and here is what we have learned.

The customer who leaves does not tell you

A lost customer does not show. There is no complaint, no empty table with their name on it. They simply stop coming, and the place is still full.

That is why almost nobody works on it: the problem is invisible. And yet it is your most profitable group, because winning back someone who already knows you costs far less than convincing someone new.

The good news is that if you have a loyalty programme or a customer database, you already have the data. You just need to look at it. If you have not set one up yet, the full guide is in what loyalty points are.

How many days to wait

This is the only decision that really matters, and it depends on your business. There is no universal number: look at how often your regular customer comes in and wait a bit longer than that.

The table below does not come from any study. It is trade judgement, meant as a starting point that you then adjust with your own data, which you will have once you have been recording visits for a few months.

The usual mistake is waiting too long. If your customer came every week and you write to them after three months, they already have somewhere else.

What to say to them

The temptation is to send a discount. Sometimes it works and it is almost always the first thing people try, but it is worth knowing what it costs.

A discount wins people back and teaches them something bad: that if they stop coming, they get an offer. Do it every month and you end up training your customers to disappear.

Alternatives that usually work just as well without touching your margin:

Tell them something new. A dish that has joined the menu since their last visit. This is the one that ages best.

Give them something money cannot buy. Trying the new menu before anyone else, booking a table that is not usually free.

Ask. A short message saying you have not seen them for a while and asking if anything went wrong. Few people reply, and the ones who do are worth a lot.

And if you do go with a discount, give it a short expiry date, so it is a nudge and not a permanent condition. In Waitry marketing vouchers are created with an expiry date, so the nudge switches itself off without anyone having to remember.

Why it runs automatically and not by hand

Because by hand it does not get done. It means checking a list every week and writing to people one by one, and that lasts about two weeks in any venue.

Set up as a trigger, you configure it once and it runs forever: someone crosses the number of days, the message goes out. Nobody has to remember.

It is, by far, the best ratio of set-up effort to return in all of a restaurant's marketing.

Choosing the channel, which is where the budget goes

Email for most people, because of cost. WhatsApp for your highest-value customers, because it gets read far more and costs far more: since July 2025 Meta charges per delivered message, not per conversation, so on a long list the difference really shows.

In other words, the same trigger can cost you little or quite a lot depending on the channel, and it is worth deciding with the estimate in front of you.

Where the list comes from

The trigger is useless if nobody is signed up, and that is the part you really need to set up first.

Contacts come from the moments when the customer is already standing still with their phone in hand: the table QR code with order and pay at the table, the kiosk screen while they build their order, your own delivery channel and the POS screen they look at while paying.

Whichever of the four they come through, they all feed the same database, which is why win-back works for dine-in and delivery customers without running two systems.

How we do it at Waitry

In Waitry it is one of the automated trigger-based sends in Waitry marketing: win back customers after X days without a visit. You set it up once and it runs on its own, just like the birthday and abandoned-basket sends.

You build the audience with your own customer filters and the tiers of your loyalty programme. You can fine-tune it: winning back someone who came every week is not the same as someone who came twice, and a top-tier customer is not the same as someone who signed up and never came back.

You choose the channel, see the estimated reach and cost before launching, and then measure the return, including what was sold after the send.

If your visit history lives in another system today, the usual answer is not to replace it but to connect it. The list of POS, payment and delivery systems we already work with is in integrations.

And if getting started on the text is the hard part, Mise drafts it for each channel for you to review.

How to know if it works

Just one thing: how many of the people who got the message came back in the following weeks.

Not opens or clicks, which tell you nothing in hospitality. Came back or did not come back.

And keep a small group with no message to compare against. Some of the people who come back would have come back anyway, and without that control group you will never know how many.

The legal side, which has a twist here

The customer must have given consent when they signed up, and every message must include an opt-out. That underlying rule is the same in every country you operate in: informed consent, a clear purpose and an easy opt-out.

In Spain two laws spell it out. The Spanish Data Protection Agency (AEPD) sums it up as processing only what each purpose needs, and article 21 of Law 34/2002 requires prior consent and a simple, free opt-out in every message.

The twist with this campaign: you are writing to someone who has not shown up for months. If they signed up back then and have not unsubscribed, their consent still stands. But if they do not reply, leave it there. Pushing someone who no longer wants to hear from you is exactly what ends in a complaint.

Shall we set it up with your data?

If you already have registered customers, a 15-minute demo shows live how many have gone X days without a visit and what it would cost to write to them. It is not a presentation: it is your own list on screen.

→ Book a 15-minute demo

And if you would rather see first how others have handled it, here are our success stories.

Frequently asked questions

How many days have to pass before a customer counts as lost?

It depends on how often people come to your venue. The rule of thumb is a little more than twice the time your regular customer takes to come back.

Do I have to send a discount?

Not always, and it is best not to make a habit of it. Telling them something new or giving access to something money cannot buy usually works just as well without touching your margin.

Which channel?

Email for most people, because of cost. WhatsApp for your highest-value customers, because it gets read far more and, since the 2025 pricing change, every delivered message is charged.

Do I need a points programme to do this?

You need to know who each customer is and when they last came. A loyalty programme is the simplest way to have that.

What if my customers are in another system?

You do not need to replace it. It connects, and the history stays where it is.

How often can I write to someone who does not come back?

Once, and if they do not reply, leave it for a good while. Pushing wears your list out.

How do I know if it worked?

By measuring how many came back, not how many opened. And by keeping a small group with no message to compare against.

Is it legal to write to someone who has not come in for months?

Yes, if they gave consent when they signed up and have not unsubscribed. And every message must include an opt-out.

Where to start

→ Create my free account

→ Book a 15-minute demo

Sources

Consent, data minimisation and opt-out in commercial messages: Spanish Data Protection Agency (AEPD) and article 21 of Law 34/2002.

WhatsApp charging per delivered message since 1 July 2025: Meta's WhatsApp Business Platform pricing documentation.

Win-back trigger, audiences by filters and tiers, cost estimate and return measurement: Waitry product documentation.

The frequencies in the table are trade judgement, not a study, and the article says so before the table.