A restaurant loyalty program works best when it's simple: use the one in your POS if there is one, reward visits or dollars with a single rule, and pick a reward that's worth more to the guest than it costs you. Most independents don't need an app of their own. Guests are open to it: in the National Restaurant Association's 2024 report, 81% of consumers not in a program said they'd join one at a favorite restaurant. Whether it pays depends on your rules.
This guide covers which kind to choose, what it costs, how to set it up and how to tell if it's working. We don't sell loyalty software and have no stake in which you pick.
Do you need one at all?
A loyalty program is worth running if you have guests who could come more often and you can get their details at the register. It's less useful if you're a destination restaurant where people come twice a year, or if you're already full on the nights it would bring people in. A rewards card won't fix a thin Google profile or a bad first visit, so check those first in the marketing guide.
The app-or-card question depends on who your guests are. Among current members, 60% prefer an app. But among people not yet in a program, 39% prefer a card and 33% an app, and 51% of baby boomers prefer a card. So don't assume you need an app. A phone number at the register suits most guests.
Which kind to run
| Type | How it works | Good for | Watch out for |
|---|---|---|---|
| Visit-based | Every visit counts: 10 visits, free entree | Coffee, lunch, regular low-ticket visits | Guests split checks to earn more; a $4 visit counts as much as $40 |
| Spend-based | Points per dollar, reward at a threshold | Full service and varied check sizes | Harder to explain at the table |
| Item-based | Buy nine, the tenth is free, or points on chosen items | Pushing a high-margin item | Can feel narrow |
| Paper punch card | Stamp it at the register | Trying the idea for almost nothing | No guest data, easy to lose, easy to fake |
| Birthday only | One reward once a year | The simplest way to collect contact details | Doesn't change how often anyone visits |
If you can only choose one, use spend-based for full service and visit-based for quick service. Toast and Square both support visits or spend.
Your options for running it
| Option | What's published, read 30 Oct 2026 | Fits |
|---|---|---|
| Square Loyalty | In Square Plus, $49 a month per location, with a 30-day trial | Square restaurants |
| Toast Loyalty | Its support articles don't give a price, and we couldn't read its product pages | Toast restaurants |
| Owner | Loyalty on both plans, $249 or $499 a month, with ordering and an app | Restaurants choosing Owner for ordering anyway |
| Thanx | No list price; quoted on restaurant count, unit volume and features | Multi-unit brands with a marketing team |
| Paper cards | Printing only | Testing whether guests care |
Start with your POS. Everything happens at the same screen, so staff are likelier to do it. On Toast, guests can sign up at checkout, at a kiosk, online or from a receipt, with just an email or a phone number. Square signs them up at the point of sale, on your site or by invoice, and counts cash purchases too.
What a reward really costs
Owners tend to price a reward at menu price and decide they can't afford it. Price it at what it costs you instead. Here's an example, with made-up inputs you should replace with your own:
| Example: spend to earn the reward | $100 |
| Example: the reward's menu price | $10 |
| Example: food cost on that dish | 30% |
| Cost to you of the reward | $3.00 |
That example costs about 3% of the spend it rewards, before labor. It looks cheap, and it would be if every point were earned on a visit that wouldn't otherwise have happened. It isn't. Your best regulars will collect rewards on visits they'd have made anyway. So the real test is whether extra visits outweigh rewards on visits you'd have had anyway, and you can only find that out by measuring. Use our food cost calculator to work out the cost of the dish you plan to give away.
Choosing the reward
- Give food or drink, not a percentage off. A free dessert costs you food cost. Fifteen percent off costs you margin on the whole check.
- Pick something they'd actually want, with a good margin: a house dessert, a drink, an appetizer.
- Make the first reward reachable. A reward many visits away is easy to ignore, so set the first one within a few visits.
- Avoid discounts on your busiest nights. Bonus points on slow nights make more sense. Square, for one, supports bonus-point promotions.
How to launch it
- 1Decide the one jobMore slow-night visits, more direct orders or more guest details.
- 2Check your POSUse what's included if you can, so there's no second system.
- 3Write one ruleVisits or dollars, and a sentence a new server can repeat.
- 4Cost the rewardAt food cost, and pick something guests want.
- 5Train the teamAsk every guest, at every check. Show the weekly count.
- 6Check at 90 daysCompare with your baseline, then change it or close it.
The steps that decide whether it works are staff ones. A sign at the register gets few signups. A server asking "are you in our rewards?" at every check gets more. Give the team a one-line script and show them the count each week.
Then use what you collect. A loyalty list is most valuable when you can reach it, but a signup isn't automatically permission to market. Text messages need written consent, so collect it separately and clearly. Our email and text guide covers the rules. If you take online orders, make the reward work only on your own ordering, and guests have a reason to leave the apps.
How to tell if it's working
Vendors publish big numbers. Square, for example, says that food and drink customers enrolled in its program spend 46% more and visit 57% more often, based on its own 2022 data. Treat that with care: guests who join a loyalty program are often your keenest guests already, so a gap between members and non-members doesn't prove the program caused it. Measure your own restaurant instead.
| Measure | How | What it tells you |
|---|---|---|
| Signup rate | New members ÷ transactions | Whether staff are asking |
| Repeat rate | Members who came back within 60 days, before and after | Whether behavior changed |
| Reward cost | Rewards redeemed × cost to make them | What you're really paying |
| Direct-order share | Direct orders ÷ all online orders | Whether guests are leaving the apps |
| Average check | Members vs your overall check, and how it changes | Whether it's growing or just discounting |
Give it 90 days. If the repeat rate hasn't moved and rewards are going to the same regulars, change the rule or close it down. Give guests fair notice and let them use what they've earned.
Common mistakes
| Mistake | What happens | The fix |
|---|---|---|
| A separate app guests must download | Few sign up | Phone number at the register |
| Rules that need explaining | Staff stop mentioning it | One sentence, or simplify it |
| Rewards as a percentage off | Margin falls on big checks | A set dish or drink |
| Texting everyone who joined | Complaints and legal risk | Separate consent for texts |
| No baseline | You can't tell if it worked | Note your numbers before launch |
| Launching and forgetting | Signups stop within weeks | Show staff the weekly count |
What to do this week
Ask your POS provider whether loyalty is included in your plan and what it adds. Pull your repeat rate if your POS shows it. Then cost one reward, and write the rule in one sentence. If you can't, it's too complicated. The free restaurant audit can show you where your guests are leaking away before you spend anything on rewards.
