Every restaurant has regulars: the office group that orders lunch three times a week, the family that comes every Sunday, the student who drops in for chai and paratha. A loyalty program gives those people a reason to choose you one more time a month. Done badly, it becomes a permanent discount that eats your margin.
This guide is about designing the program so it pays for itself.
Choose a format
| Format | How it works | Good for | Watch out for |
|---|---|---|---|
| Stamp card | Buy 9 chai, get the 10th free | Cafés, chai dhabas, bakeries | Cards get lost or faked |
| Points | Earn points per rupee spent, redeem for rewards | Restaurants with varied bill sizes | Needs a system to track balances |
| Member deals | Special prices on set days for members | Filling slow days | Can train customers to only come on deal days |
| Birthday or anniversary reward | Free dessert or discount on the occasion | Family restaurants | Needs correct dates and reminders |
For most restaurants with bills ranging from Rs 500 to Rs 8,000, a points program is fairer than stamps, because a family spending Rs 6,000 earns more than a single customer spending Rs 600. Stamp cards still suit a café where nearly every order is a similar size.
Set the reward rate
The reward rate is how much value you give back for each rupee spent. Keep it modest and simple to explain. A common structure:
- Earn 1 point for every Rs 100 spent.
- Redeem 150 points for a free item priced at Rs 500, such as a dessert or a starter.
To earn 150 points, a customer spends Rs 15,000. They get Rs 500 of menu value back, which is 500 ÷ 15,000 = about 3.3% of their spend. That feels real to the customer without hurting you, and the real cost to you is even lower, because the free item costs you only its ingredients.
Do the maths before you launch
Here is a worked example for a restaurant with 200 active members. Each member visits twice a month and spends Rs 1,500 per visit, so 400 member visits a month before the program.
Suppose the program gets 50 of those members, a quarter, to add one extra visit a month.
| Line | Calculation | Amount |
|---|---|---|
| Extra visits | 50 members × 1 visit | 50 visits |
| Extra sales | 50 × Rs 1,500 | Rs 75,000 |
| Gross profit on extra sales (food cost 35%) | Rs 75,000 × 65% | Rs 48,750 |
| Total member visits after | 400 + 50 | 450 visits |
| Points earned per month | 450 × 15 points | 6,750 points |
| Rewards earned | 6,750 ÷ 150 | 45 rewards |
| Cost of rewards (food cost Rs 175 each) | 45 × Rs 175 | Rs 7,875 |
| Monthly gain | Rs 48,750 − Rs 7,875 | Rs 40,875 |
Two things stand out. First, the reward cost is the food cost of the free dessert (Rs 175), which is far below its menu price of Rs 500. Second, the program only pays off if it changes behaviour. If no member adds a visit, you are simply giving away Rs 7,875 a month. So measure visits per member before and after launch.
If food cost is not yet clear to you, read food cost percentage first; the whole calculation depends on it.
Pick the right rewards
- Choose rewards with low food cost and high perceived value: desserts, drinks, fries, a naan basket.
- Avoid giving your expensive stars, such as mutton karahi, as rewards.
- Use rewards to introduce dishes you want to sell more of. If a new handi is struggling, make a half handi a reward and let customers try it. Our menu engineering guide explains which dishes to push.
- Set a validity period, such as points expiring after six months, and tell customers clearly.
Use WhatsApp, with respect
In Pakistan, WhatsApp is where your customers already are. It is the natural place to send a points balance, a birthday offer or a weekend deal. A few rules keep it useful instead of annoying:
- Ask permission when you sign the customer up. Write down that they agreed.
- Send at most one or two promotional messages a week.
- Make every message specific: '120 points, 30 more for a free kulfi' beats 'Visit us today!'
- Time offers for your slow periods, such as weekday lunch, instead of Friday night when you are full anyway.
- Stop messaging anyone who asks you to.
Mistakes that sink loyalty programs
- Rewards that are too far away. If a customer needs to spend Rs 50,000 before getting anything, most will forget about the program long before.
- Rewards that are too close. Giving 10% back on every bill is just a permanent discount, and it pulls your margin down on customers who would have come anyway.
- Rules the cashier cannot explain. If the staff do not understand the program, they will not mention it, and customers will not join.
- No end date on points. Old unredeemed points sit on your books as a promise you have to keep, and customers who return after two years expect them to work.
- Stacking with deals. Decide whether points can be earned on deal orders and redeemed on top of a deal. If both are allowed, your cheapest-margin orders get even cheaper.
Cafés and bakeries: a stamp card example
A café selling cappuccino at Rs 550 with a food and cup cost of Rs 160 can run a simple 'buy 9, get the 10th free' card. Across ten cups, the café collects 9 × Rs 550 = Rs 4,950 and its cost is 10 × Rs 160 = Rs 1,600. The free cup reduces the margin per cup from Rs 390 to (4,950 − 1,600) ÷ 10 = Rs 335, which is a small price for a customer who keeps coming back.
Sign-up should take ten seconds
The cashier asks for a mobile number at billing, enters it, and the points are added to that number. Nothing more. Long forms at a busy counter kill sign-ups. You can collect a name and birthday later, through a WhatsApp message.
Tracking it properly
Points programs need a system that ties points to a phone number and to every bill, otherwise balances become an argument at the counter. ArfaPOS includes loyalty points, time-limited deals and WhatsApp customer marketing in the same system as billing, which is covered on the restaurant loyalty program page. Whatever you use, check these numbers every month:
- Number of active members (visited in the last 60 days).
- Average visits per member per month.
- Average bill of members compared with non-members.
- Rewards redeemed and their food cost.
Run the program for three months before judging it. The first month is mostly sign-ups; the extra visits show up after customers start seeing their points grow.