Most cafés already run a loyalty programme. It is a piece of card the size of a business card, it lives in the guest's wallet behind a bank card, and right now it has six of ten squares stamped. The thing a digital loyalty card has to beat is not another app. It is that.
The paper card is not a bad product. It costs almost nothing, staff learn it in a second, and no guest has ever needed it explained. Any honest case for replacing it starts there. What follows is the arithmetic on one ten-stamp card, and then the thing the card has never been able to tell you.
What a free coffee actually costs
Take a $5 flat white in a café running a 75% gross margin on drinks. Beans, milk, cup and lid come to $1.25. Buy nine, get the tenth free. Substitute your own cup price and your own margin as you read; the shape of the answer does not change.
Across one full card the guest pays $45 and drinks ten coffees. Your cost of goods is $12.50 either way, because you make ten drinks either way. The card leaves you $32.50 of gross profit where ten paid drinks would have left $37.50. That is a 10% discount on the ten coffees and a 13% haircut on the gross profit for the cycle.
Most operators stop there, decide 10% is affordable, and carry on. The more useful number is underneath it, because that free coffee has two possible costs and they are not close to each other.
If the guest was coming in anyway, the free drink cost you $5.00. You handed back a sale you already had. Nothing about the reward changed anything they did.
If the reward is what brought them through the door, the free drink cost you $1.25 in goods and bought a visit you would not otherwise have had.
Same cup of coffee, four times the cost depending on who is holding the card. Every stamp card in circulation is some mix of the two, and the programme is a good one or a bad one entirely according to that mix. The paper card cannot tell you what your mix is.
The real cost of the paper card is that it is anonymous
A punch card is a bearer instrument. Whoever holds it gets the coffee. It carries no name, no history, and no record of anything that ever happened to it.
So a set of very ordinary questions have no answers. How many cards are out there right now? How many got to three stamps and were abandoned in a coat pocket? Is the guest redeeming this morning a five-day-a-week regular or someone on their fourth visit this year? Did the guest who filled a card in March come back in April? None of it was written down, so none of it can be answered.
That is the cost that never appears anywhere. You cannot price the programme, so you cannot improve it. You cannot tell a reward that worked from one that was a gift. And you cannot do the thing loyalty is supposed to make possible, which is speak to the guest. A regular can stop coming entirely and the card will never mention it.
What the café learns that the card never told it
A digital loyalty card is the same mechanic with a record attached. Nine stamps, tenth free, still the deal. The difference is that every stamp is now a row in a table with a guest on the end of it, which turns four unanswerable questions into reporting:
- Who. A name and a way to reach them, captured once at the counter instead of never.
- How often. Visits per month per guest, which is the number that separates the regular from the tourist and decides which of your two free-coffee costs you are actually paying.
- What they order. The oat flat white every morning, the pastry only on Saturdays. Useful for the roster, for the bake order, and for knowing what a reward should be.
- Who stopped coming. The one paper could never do. A guest who came four times a week for a year and has not been in for three weeks is a visible event rather than a face you eventually stop picturing.
None of this is new to your competition. The large branded chains have run app-based loyalty for years. Independent cafés were never behind on the coffee. They were behind on knowing who came back.
What it replaces day to day
Less than operators expect, and more than the card. The stamp still happens at the counter and still takes a second. What goes away is the small tax around it:
- The reprint run every few months, and the box of blanks behind the till.
- "I left it at home." "I lost it." "Can you just stamp two?"
- A rubber stamp that anyone can buy online for the price of a coffee, and the quiet fraud that follows it.
- A card stamped at your first site meaning nothing at your second one.
Habitu builds this as the café's own branded app rather than a shared directory app, so the card carries your name and not a platform's. Visits come off the till: Square and Foodics are the point-of-sale integrations that ship today. Pricing runs $49 to $499 per location per month, with a per-order charge on the lower plans. The app ships in English and Arabic. Pretzel Australia runs its guest app on Habitu, and there are operators on the platform in Australia, Canada and the US.
When the paper card is still the right answer
One site, one till, an owner on the floor most days who knows the regulars by name. The informal layer is doing the job software does at scale, and the card is a token of a relationship that already exists. Paper is fine there. It is cheap, and it does not mislead you, because it never claimed to tell you anything.
The threshold is the morning you stop recognising faces. A second site. A third shift. Enough volume that the 8am queue is a blur. At that point the relationship you were relying on is gone and nothing has replaced it, and the free coffees keep going out the door at a cost you have no way of working out.
So the test is not "should we go digital". It is narrower and harder to dodge: of the free coffees you gave away last month, how many bought you a visit you would not otherwise have had? If you cannot answer that, the problem is not that the programme is failing. It is that you have no way of knowing either way, and the card is what is keeping you there.