A loyalty programme does not create regulars: it rewards the ones you already have. That is useful, but it means the first question is not “which mechanism” but “what does the discount cost me when I give it to someone who was coming back anyway”. Here are the three mechanisms, their real cost, and the common case where a customer record is enough.
In short
- The discount is a cost, not a marketing budget. It comes out of your margin, dish by dish.
- The stamp card is the simplest and the most expensive in margin if the tenth free item is a dish, not a coffee.
- Points only help if you read them. With no report it is a discount dressed as a programme.
- The customer record returns most: it costs nothing and tells you who comes back.
- Measure before launching. With no baseline frequency you will never know whether it worked.
The three mechanisms, and what they cost
The cost is not the price of the printed card, it is the margin you give back. Read the third column before the first.
| The mechanism | What it needs | What it really costs |
|---|---|---|
| The stamp card | Cards, a stamp, and nothing else. | The whole margin of the free item, and you do not know who got it. |
| Points | A till that counts them and a customer record. | A percentage on everything, including on already-loyal customers. |
| The customer record | One record per customer, created at the till in ten seconds. | Nothing. It gives no discount, it gives information. |
| The credit ledger | A balance tracked per customer, settled at month end. | Nothing in margin, but risk: it is cash you have lent. |
The last two rows are not loyalty programmes in the marketing sense, and that is exactly why they work: they build loyalty through service and recognition, without giving margin back.
The calculation nobody does before launching
Three numbers are enough, and they are already in your till.
- The margin of the free item: A free coffee on the tenth costs its ingredients, a few dirhams. A free dish on the tenth costs the margin of a dish. They are not the same programme, and many cards do not distinguish.
- The share of already-loyal customers: If three quarters of your stamps go to people who came weekly anyway, you have cut your prices for them and gained nothing. That is the programme’s invisible cost.
- Baseline frequency: How often a regular comes back, before the programme. Without that figure you will never be able to say frequency rose, and you will conclude from an impression.
That last point is the one most often skipped, and it is what makes the programme measurable. It is read from your till’s history if sales are attached to customer records, which is precisely the cheapest of the four mechanisms.
How to launch without getting it wrong
Five steps, two of them before printing anything.
- Create customer records for a month, with no reward at all. Attach the sales. You are measuring.
- Read the real frequency: how many customers come back, at what rhythm, with what basket.
- Choose the reward by margin, not by what feels generous: giving the drink costs little, giving the dish costs the dish.
- Set a limit: an expiry, one benefit at a time, and a cap. A programme with no limit is a permanent discount.
- Reread the frequency three months later, on the same customers. If it has not moved, stop the programme: it is costing margin and changing nothing.
Step 5 is the one almost nobody dares do. A programme that is not measured never ends, and that is how a permanent discount settles into a menu without anyone deciding it.
The credit ledger is not a loyalty programme
In many Moroccan shops real loyalty runs through the tab: the customer settles at month end and comes back because they have an account with you. It is effective, and it is not free: it is cash you are advancing, with a risk of non-payment. Treat it as such, with a balance tracked per customer and a ceiling, not as an informal favour held in someone’s head. Tracking makes credit sustainable; the absence of tracking makes it expensive.
What works in Morocco, and what works less
What works: recognition. A waiter who knows the usual order, an up-to-date balance, a one-off gesture decided on the spot. That needs no programme, only a customer record visible at the till while you serve, and it is often worth more than a tenth free coffee three weeks later.
What works less: loyalty apps that ask the customer to install something. In a neighbourhood business the adoption rate is low and the cardboard card does better. And message reminders work mainly when they carry useful information, something new or an opening time, rather than regular solicitation that ends in a block.
Signs a programme is doing nothing
If two of these five are true, stop it.
- You do not know how many cards are in circulation.
- You never measured frequency before launching.
- The same faces fill their cards and they were your regulars already.
- The reward is a dish while your margin is in the drinks.
- The programme has no end and no cap.
Mistakes to avoid
- Launching before measuring. With no baseline, no review is possible.
- Giving away the highest-margin product. It is the dearest for you and often not the most wanted.
- Confusing discount with loyalty. A permanent discount attaches nobody, it simply lowers your prices.
- Keeping credit in your head. With no tracked balance, cash evaporates politely.
- Never stopping a programme. The cost does not stop either.
Frequently asked questions
Do loyalty programmes really bring customers back?
They mainly reward the ones who were already coming back. That is useful for maintaining a relationship, but it does not create frequency on its own. The only way to know in your case is to measure frequency before launch and again three months later, on the same customers.
Stamp card or points: which should I choose?
The stamp card is simpler and enough for a business with a regular basket; points only matter if your till counts them and you read the report. Either way the choice of reward counts more than the mechanism: give away what has little margin, not what has the most.
How much does a loyalty programme cost?
Not the price of the cards, but the margin given back. Multiply the margin of the free item by the number of rewards handed out, then remove the share that went to already-regular customers: what is left is the real cost of buying extra frequency.
Is a credit ledger a good idea?
It is an effective loyalty mechanism in a neighbourhood business, and it is also advanced cash. It holds up if it is tracked: a balance per customer, visible at the till, with a ceiling. Kept from memory or in a notebook it becomes a slow loss nobody can date.
Do I need an app to build loyalty?
Rarely in a neighbourhood business: asking a customer to install something excludes most of them. A customer record in your till, with history and balance, gives you the substance without asking the customer for anything.
What to take away
Start with the customer record, not the programme. A month of sales attached to customers will tell you who comes back, how often and with what basket, and it costs not one dirham of margin. If, with those figures in hand, a programme is justified, choose the reward by its margin, give it a limit, and put a date in three months to decide whether to keep it or stop it.
Start by knowing who comes back
BelloPOS keeps a record per customer with their purchase history and balance, visible at the till while you serve. It is included in the Lite licence, free for life and entirely offline.
Read next
Other practical guides on the same subject: