How to launch a loyalty program for a Montréal café in one week
6 min read
A café loyalty program fails in one of two places. Either the math is wrong, so every reward eats margin you didn't have, or the counter is wrong, so your staff quietly stop offering it by Thursday. Both problems are avoidable, and neither takes months to solve. Here is a one-week plan for a Montréal café or dessert counter, day by day, from the reward math to the checkpoints that tell you whether it worked.
The plan assumes you're launching on Perkaria — a tablet beside your register, a wallet card on the customer's phone, nothing to install on your POS — but most of it applies to any program you'd run at a busy counter.
Day 1: get the reward math right
Start with a stamp card, not points. It's the mechanic your customers already understand and the one your staff can explain in five seconds mid-rush. Points per dollar are worth adding later, once enrolment is a habit; on the loyalty side of Perkaria that's a settings change, not a relaunch.
Then price the reward honestly. Say you run a 10-stamp card and the tenth drink is free. If that drink sells for $5 and costs you about $1.50 to make — your number will be different, use yours — then nine paid visits fund each reward, and the card costs you roughly 17 cents a visit ($1.50 ÷ 9). That's the real question for any reward: not what it sells for, but what it costs you, spread over the visits it takes to earn it.
- Pick a goal a regular can actually reach in four to six weeks. Ten stamps suits a coffee habit; six suits a weekly dessert stop.
- Make the reward something with good margin and real appeal: a drink or a slice, not a percentage off the whole bill.
- Decide where double stamps go. Aim them at your flattest hours — weekday mid-afternoons, say — so the program buys visits you weren't already getting.
Day 2: set the guardrails while it's calm
Launch week is a bad time to discover your program leaks. Set the daily earn cap per member, so one generous afternoon can't inflate a balance, and decide who gets a staff PIN — every stamp and redemption on the counter tablet is attributed to the person who did it. One thing you don't have to configure is verification: every stamp lands as pending and is checked overnight against the daily sales export from your register. No integration, no API. If your POS can export a sales file, you're compatible. How the whole loop works is worth ten minutes of your Tuesday.
Day 3: set up the counter
The tablet lives where the payment happens, not at the end of the bar. If customers tap their card at the register, that's where enrolment and stamps have to happen too, or they'll get skipped the first time there's a line.
- Put the tablet within arm's reach of the register, plugged in, screen always on.
- Run five test enrolments with staff phones, timer out. If it takes longer than the debit terminal, rearrange until it doesn't.
- Check what the customer will see: your logo and colours on the wallet card, and the French and English wording your members get.
- Pick your launch day: a medium-busy one. Not your dead Monday (not enough practice), not Saturday (not enough patience).
Day 4: brief the staff, with an actual script
Your program will be exactly as good as the sentence your staff say at the counter. Don't leave that sentence to improvisation. Write it down, tape it near the register for the first week, and let people bend it to their own voice once they've said it fifty times.
- Enrolment is a staff action: you ask, you type the number, the customer gets a magic link by text that opens their card. There's no QR code on the table for self-serve sign-up, and that's deliberate — consent gets asked out loud, by a person.
- Stamps show as pending right away and get verified overnight against the sales export. Nobody stands at the register waiting.
- If someone doesn't want texts, that's a no. Consent is the whole game, and STOP is honoured instantly.
Days 5 to 7: the launch rhythm
A launch is not a poster; it's a habit. For the first three days, the only goal is that the offer gets made — to every customer paying at the counter, not just the friendly ones.
- Set a shift target for offers made, not sign-ups. Your staff control the ask, not the answer.
- Hold a two-minute debrief at close: which line of the script feels awkward? Fix the script, not the person.
- Enrol your own regulars yourself. The people you know by name should hear about it from you.
- Watch enrolments per day in the dashboard, and expect a dip after day one. That's normal — the habit hasn't formed yet.
The first month: three checkpoints
After launch week, the work moves from the counter to the owner dashboard. Skip the vanity numbers; three things tell you whether the program is working.
- Week 2 — enrolment pace. Is the counter still making the offer? If sign-ups fell off a cliff, the problem is the script or the ask, not your customers.
- Week 3 — visits per member per month. This is the number the program exists to move. Note your baseline now and judge every future change against it.
- Week 4 — your top customers and the lapsed list. You'll recognize names. The moment that list stops being abstract is the moment the tool starts earning its keep.
What to skip in week one
Two temptations to resist. First, don't launch stamps, points and tiers all at once; one mechanic explained well beats three explained badly. Second, don't send a marketing blast in week one — build the list before you use it. When you do start campaigns, the guardrails are enforced by the product rather than by your good intentions: quiet hours from 9 p.m. to 9 a.m., a hard cap of two marketing texts per member per week, and opt-outs honoured on the spot.
One week is enough. The math is a morning's work, the setup is an afternoon, and the rest is a sentence said at the counter until it becomes a reflex. The cafés that win at this don't have the cleverest rewards — they're the ones where the ask happens every time.