Restaurant Loyalty Programs: What Actually Works

Guests belong to four or more restaurant loyalty programmes on average. They actively use 1.5 of them. That is from Simon-Kucher's study of over 500 restaurant customers and 100 industry experts, and it means the average scheme is one of the two and a half that get ignored.

Which is the problem with starting from the reward. A stamp card is easy to print and easy to forget, and the guests who fill one in are usually the guests who were coming back anyway.

Here is what a returning guest is actually worth, why most schemes miss them, and what to do instead if you run a small dining room rather than a chain.

Key Takeaways

• Guests belong to four or more restaurant loyalty programmes and actively use 1.5 of them, so the average scheme is one of the ones being ignored (Simon-Kucher).
• A 10 percent discount handed to 100 guests a month costs 400 euros, and if 80 of them were returning anyway, 320 euros of that buys behaviour you already had.
• Ten guests who return four times a year are 1,600 euros of revenue carrying no booking fee, ad spend or commission.
• The lever is recognition, not reward. Knowing a name is on its second booking costs nothing and outperforms a stamp card.
• Skip a loyalty programme entirely if you are tourist-heavy, naturally low-frequency, or have not yet fixed service and the online menu.

What a returning guest is worth

Do the arithmetic on your own numbers rather than trusting a percentage from someone else's market.

Take a venue with 150 bookings a month and an average bill of 40 euros. A guest who comes back four times in a year is 160 euros of revenue. Ten such guests are 1,600 euros, and not one of those visits costs you a booking fee, an ad or a commission. You already paid to meet them the first time.

Now price the usual reward against that. A 10 percent discount on a 40 euro bill costs you 4 euros. Give it to 100 guests a month and you have spent 400 euros. If 80 of them were coming anyway, 320 euros of that went to reward behaviour you already had, and 80 euros went to change behaviour you did not.

That ratio is the whole argument. It is not that discounts fail, it is that untargeted discounts mostly land on people who need no persuading.

Why most schemes get ignored

The same study found 70 percent of members join specifically for financial benefits, so a scheme built purely on money attracts people shopping on price, and keeps them only while the offer runs.

Meanwhile the thing that actually separates a regular from a stranger is duller. Being recognised. The host knowing it is a second visit, that last time was a birthday, that one of the two does not eat fish.

None of that needs a points engine. It needs a record.

What brings people back

Four things, in ascending order of effort. The first two happen in the dining room and need no software at all.

1. Spend the discount on one table instead of all of them. The arithmetic above spread 400 euros across 100 guests and changed almost nobody's mind. Take a small part of that and put it on the table that clearly had a good evening: the last round, the dessert, a coffee that never reaches the bill. It reads as generosity rather than as a promotion, it costs a few euros instead of four euros a head across the room, and the guest remembers who did it. This is the only item here that costs money, and it still costs less than the discount it replaces.

2. Write down one thing about them. Not a profile. One line on the booking: anniversary in March, does not eat shellfish, asked for the corner table. Next time it sits in front of whoever greets them, and being remembered is the whole product here. This is the cheapest thing in the article and the one that separates a room people like from a room people belong to.

3. Know that it is a second visit. Most venues cannot tell. The booking arrives, the table turns, the name disappears. If your reservation records survive past the shift, a repeat name is visible without any extra tooling, and points 1 and 2 stop depending on whoever happens to be working that night.

4. Send one letter a month, if you will actually keep it up. This is the one that takes real discipline, and most kitchens never get there. That is fine, the first three are worth more. A newsletter is easier to sustain than personal notes, because you write it once and it goes to everyone.

The address needs handling before you send anything. EU rules let a business email its own past customers about its own similar offers, as long as you give them a way out both when you take the address and in every message (ePrivacy Directive, Article 13). Countries apply that exception differently, so the safe version is an empty tick box at the booking that the guest ticks themselves. A pre-ticked one has never counted as consent (Planet49, Court of Justice of the EU).

Once a month is the ceiling, and the test before sending is whether you can say in one sentence what the reader learns: the autumn menu is on, the terrace opens Friday, you close for two weeks in July.

The autumn menu starts on the 12th. Game, mushrooms, and the beetroot dish people kept asking for last year. Thursdays are the quiet nights if you want the corner table.

If the honest answer is that you want covers on Thursday, that is not news, it is an advert. And a letter without news teaches people to ignore the next one.

Every reservation you take already carries a name, a date and a way to reach them. Hollapp keeps all of it in one place, attached to the guest rather than to the single booking: the visit history, the note about the corner table and the shellfish, the address you need for the newsletter. A second visit shows up as a second visit, so all four points above run off the reservations you are already taking, not off a separate loyalty tool you have to buy and keep running alongside them.

A host who remembers the corner table beats a card with nine stamps on it.

When not to run one

Skip it, honestly, in three cases.

If you are in a tourist position where most guests are in the city for three days, they are not coming back no matter what you print. Your lever is the review they leave, not the discount you offer.

If your frequency is naturally low, the maths does not carry. A destination restaurant people visit for anniversaries has a two-visit-a-year ceiling, and a points scheme will not move it.

And if the basics are not right, fix those first. A loyalty programme layered onto slow service or an out-of-date menu online just buys you a discounted version of the same problem. Reward schemes amplify what you already have.

Common questions

Do restaurant loyalty programmes actually work?

Some do. The problem is that most start from a discount, and a discount reaches everyone equally, including the guests who needed no persuading. Simon-Kucher found 70 percent of members join for financial benefits, so a money-only scheme attracts price shoppers and holds them exactly as long as the offer runs. The ones that work start from recognition.

How much is a returning guest worth to a restaurant?

Run it on your own numbers. At a 40 euro average bill, a guest returning four times a year is 160 euros, and ten such guests are 1,600 euros carrying no booking fee, ad spend or commission. Compare that against what an untargeted discount costs you over the same year before you decide a scheme is worth running.

Do I need loyalty software to run a loyalty programme?

No. The three cheapest tactics here need nothing beyond your reservation records: put the discount on one table rather than the whole room, write one line about the guest on the booking, and be able to see that a name is on its second visit. Software matters when those records have to survive a staff change.

Can I email guests who booked a table with me?

Usually yes, within limits. EU rules let a business email its own past customers about its own similar offers, provided an opt-out is offered when the address is taken and in every message (ePrivacy Directive, Article 13). Countries apply that exception differently, so the safe route is an empty tick box at the booking. A pre-ticked box has never been valid consent.

When should a restaurant not run a loyalty programme?

In three cases: a tourist position where most guests are in the city for three days, a naturally low-frequency venue such as an anniversary restaurant with a two-visit-a-year ceiling, and any venue where service or the online menu is still wrong. Rewards amplify what is already there.

Where to start

You do not need software you do not own yet. Start with what the bookings already give you.

Collect the name and contact once, at the booking, and keep it. Then take one table a night, the one that clearly had a good evening, and send something small over that never reaches the bill. That is the whole first week.

After that, look at last month and count how many names appear twice, because that number is your actual retention rate and most operators have never seen it.

If the number of repeat names is already healthy, you have a recognition problem rather than a loyalty problem, and the fix is in the dining room. If it is close to zero, that is worth knowing before you spend anything on rewards.

Guests you already have are the cheapest covers you will ever fill, which is the same argument as what a reservation system really costs: the audience you own beats the one you rent. If you want the guest history kept for you, it comes with reservations on every plan, including the free one.