Loyalty Program ROI: The Math of a Repeat Customer for a Coffee Shop

The variables that actually determine ROI

Coffee shop loyalty ROI comes down to four numbers, and none of them are mysterious. First, your average ticket. Second, your gross margin on that ticket - coffee has famously good margins, which is why a free-drink reward is cheaper than it looks. Third, the reward cost itself: what a completed card costs you at wholesale, not retail. Fourth, and most important, the behavioral lift: how many extra visits the program creates that would not have happened otherwise. Everything else - signage, staff scripts, the platform fee - is small by comparison.

• Average ticket per visit (use your POS report, not a guess)
• Gross margin on that ticket
• Reward cost at your wholesale/ingredient cost
• Extra visits generated: the only variable the program exists to move

A worked example: the 10-stamp coffee card

Run the numbers on a simple "buy 9, get the 10th free" card at a hypothetical shop with a 20-currency-unit average ticket. Assume ingredient cost on a redeemed drink is about 1 unit, so a completed card costs the shop 1 unit in reward cost per 180 units of revenue earned across the ten visits - roughly half a percent of revenue. That is the entire direct cost of the program. Even after adding the platform subscription, the program only needs to generate a handful of extra visits a month to break even, because each incremental regular contributes margin on every visit, not just one.

How frequency shifts the math

The real ROI lever is visit frequency. If a stamp card moves a customer from two visits a week to three, the program did not add one visit - it added roughly fifty visits a year from that single person, all at coffee-shop margins. This is why the stamp mechanic outperforms flat discounts for frequency businesses: the visible progress ("three stamps to go") pulls the next visit forward, while a percentage-off coupon rewards a purchase that mostly would have happened anyway. When you model ROI, model the frequency change per member, not the discount depth.

How to measure it honestly in your own shop

Set a simple baseline before launch: average visits per week from your POS, split by new versus returning customers if your system allows it. After 60 to 90 days, compare loyalty members against non-members on visit frequency and average ticket. Two honest caveats: members may already have been your best customers (selection bias), and weather, season and local events move everything. The cleanest signal is the stamp-completion curve - how many members who earn a first stamp go on to complete a card - because that measures the habit the program was built to create.

• Baseline visits per customer before launch, from POS data
• Compare member vs non-member frequency after 60-90 days
• Watch the stamp-completion curve, not just enrollment counts
• Recalculate reward cost at wholesale, never retail

When a loyalty program is NOT worth it

The math fails in predictable situations: if the reward is so rich it erases margin on completed cards, if staff never mention the program so enrollment stalls, or if the reward is something customers do not actually want. It also underperforms for pure one-time-transaction businesses with no natural repeat cycle. But for a coffee shop - high margin, high frequency, routine purchase - a simple stamp card is one of the few marketing expenses where the arithmetic genuinely leans in your favor, and a digital card in Apple Wallet or Google Wallet removes the printing cost and the forgotten-card problem that sank paper programs.