Restaurant loyalty programs: what rewards really cost, and what pays back

The punch card is in a kitchen drawer, or it went through the wash. Your regular still comes in every Thursday, and you're about to give them a free main for something they were going to do anyway.

A restaurant loyalty programme (loyalty program in the US) pays back only when the extra visits it causes earn more margin than the rewards cost. Work out each reward's cost as a percentage of member spend, using what the reward costs you to make rather than its menu price. Then check whether members really visit more often than they did before they joined. Last checked: 4 November 2026.

This guide covers the UK and the US; where the rules differ, they are labelled.

Do restaurant loyalty programmes work?

On average, yes, but modestly, and design decides whether yours does. The independent research, as opposed to loyalty vendors' own figures:

Guests join readily. The National Restaurant Association (9 April 2024) found 52% of US consumers already take part in restaurant loyalty programs, in 3.6 programs on average. In the UK, the American Express Rewards Report 2026 reported by Retail Times (9 September 2026) found 90% of adults belong to at least one scheme, and 50% to a food and drink chain's.

What surveys can't show is cause. People who join are often your best guests already, so comparing members with non-members mostly measures who signs up. Most uplift figures you will see come from loyalty software companies making exactly that comparison.

How much eating out is there to win?

Less than before the pandemic, in both countries. Our analysis of official household spending data (4 November 2026):

Method: US households spent 17.5% more dollars eating out in 2025 than in 2019, while the BLS price index for food away from home (series CUUR0000SEFV) rose 34.4% between the two annual averages. The UK figures are ONS's own real-terms series (table 4.1), which also includes hotels; ONS changed its survey weighting during and after 2020, so read the UK figure as a direction, not a precise size. Restaurant and café meals alone came to £17.80 a week per UK household in the year to March 2025.

What it means: your loyalty scheme is competing for a share of fewer meals out. A scheme that hands discounts to regulars who would have come anyway costs more than ever, and one that wins a visit from a competitor is worth more.

Should your restaurant run a loyalty programme?

Run one only if guests come back often enough to reach a reward, and you can afford it at your margins. Use this test:

Rewards don't have to be discounts. Priority booking on a busy night or a dish before it reaches the menu costs you little. That is our reasoning, not a measured result, so test it against a conventional reward.

What does a loyalty reward really cost?

A reward's real cost is its value as a share of the spend needed to earn it, times what it costs you to give, times the share of rewards redeemed.

Reward cost as % of spend =
  (reward value ÷ spend needed to earn it)
  × cost ratio
  × share of rewards redeemed

Cost ratio = your food cost % for a free menu item (for example 30%)
           = 100% for money off the bill

Worked example. A guest earns 1 point per £1 or $1 spent, and 100 points buys 5 off. The face value is 5 ÷ 100 = 5% of spend. If the reward is a menu item at 30% food cost, its real cost is 5% × 30% = 1.5% of spend. If it's money off the bill, you give up the full 5%. If only 70% of rewards are ever redeemed, money off costs 3.5% and the menu item about 1.05%.

The food cost calculator works out your food cost per dish. Here is the same maths for four common structures, at 30% food cost and with every reward redeemed:

The visit card shows why menu value misleads. Nine mains at 18 each is 162. The free main has a menu value of 18 (11.1% of 162), but at 30% food cost it costs 5.40 to make, which is 3.3%.

In the UK, the VAT treatment of free items and discounts can change the real cost. Under revenue accounting standards (ASC 606 in the US, IFRS 15 or FRS 102 in the UK), unredeemed points can count as a liability on your books. Ask your accountant how both apply to you.

Points, visits or tiers: which suits your restaurant?

Choose visits if guests come often and spend similar amounts, points if spend varies, and tiers only if you have a large base of regulars.

Three design rules come straight from the studies above:

• Give a head start. Bonus points or stamps on joining made the café card and the car wash card more likely to be finished, and faster, for the same spend.
• Keep the first reward close. Guests sped up as the reward got near, so a reward they can see from the start does more than a big one far away.
• Restart momentum after a reward. In the café study, members slowed down again after redeeming. Starting the next card with a bonus stamp is our inference from that, worth testing.

How do you tell if your loyalty programme is paying back?

Run a break-even test before you launch, then check it after three months. Every figure below is a hypothetical, rounded assumption. Read each amount as pounds or dollars and replace it with your own.

So the scheme pays for itself if it causes about 215 visits a year that would not otherwise have happened: roughly one extra visit for one member in five. Every visit beyond that adds 28 of margin. For where that margin sits in your accounts, see restaurant profit margins in the UK or the US.

After launch:

• Compare like with like. Look at each member's visits per month before they joined and after. Don't compare members with non-members.
• Watch redemptions. If almost nobody redeems, the reward is too far away. If redemptions bunch at peak times, move rewards off-peak.
• Count only the extra visits as the programme's return, not all member sales.

What rules apply to messages you send members?

Earning points needs no marketing consent, but emailing or texting members about offers does:

• UK: marketing emails and texts fall under PECR and UK GDPR. You need consent, or the "soft opt-in" for existing customers, where you collected their details during a sale, market your own similar products and give a clear opt-out every time. See the ICO's guidance on electronic mail marketing.
• US: commercial email must follow CAN-SPAM. The FTC's compliance guide requires a clear opt-out honoured within 10 business days. Marketing texts fall under the TCPA, which generally requires prior express written consent.

Put any points expiry in the terms in plain words and tell members before points lapse. Check with an adviser for your situation.

A one-page loyalty plan you can copy

Fill in the square brackets before you launch, and keep it next to your reward cost table.

LOYALTY PLAN: [Restaurant name]

Fit check:       Most guests visit [__] times a year (need 4+)
Name:            [Name] Rewards
Who can join:    Any guest who orders and pays
Earn:            1 point per [£1 / $1] spent on food and drink
                 (not on tips, service charge or gift card purchases)
Welcome bonus:   [20] points on joining, so the first reward is close
Rewards:         [100] points = [dessert or coffee]
                 [250] points = [main course up to £/$ ___]
After a reward:  [10] bonus points to restart progress (test it)
Off-peak extra:  Double points [Monday to Wednesday]
Not valid:       [Friday and Saturday after 7pm] (optional)
Expiry:          Points expire after [12] months without a visit;
                 members are told [30] days before
Real cost:       [__]% of member spend (from the cost table)
Break-even:      [__] extra visits a year
                 = (reward cost a year + running cost) ÷ margin per visit
Measure:         Each member's visits a month, before vs after joining
Review date:     [3 months after launch]

Can TablePort handle this for you?

Yes. TablePort's loyalty programme is built into the till and is part of the plan, not a paid add-on, so the plan above runs without punch cards.

1. Set the earn rate per order, per item or per pound or dollar spent. Points post when the bill is paid, with nothing for the server to scan.
2. Add reward steps at the thresholds your cost table supports, such as a free item at 100 points.
3. Guests join the first time they order and pay, and carry a branded Apple Wallet or Google Wallet pass, so there is no app to download or card to forget.
4. Sort the member list by lifetime value, order count and points balance to see who your regulars are.

One limit: it cannot tell you which visits the programme caused. That still needs the before-and-after check above. See Loyalty.