Most specialty cafés have a loyalty program. Most loyalty programs lose money — they give away free coffees to customers who would have come back anyway, with little measurable retention benefit. The good loyalty programs do two things: they drive incremental visits, and they build a relationship that makes customers feel known.

This is what the difference looks like in practice.

The math of a punch card

The classic "buy 9, get 1 free" punch card is the most common loyalty mechanic in coffee. The math, on a $5 cortado:

Customer pays $45 for 9 cortados.
Café gives away the 10th — $5 in retail value, but only about $1 in variable cost (beans, milk, cup).
Effective discount on the full set of 10 cortados: $5 / $50 = 10%.

That 10% feels manageable. But it's only worth giving if the customer wouldn't have come the 10th time without the card. Industry surveys consistently show that 60-75% of frequent customers would have returned regardless — meaning the 10th coffee was already coming. For those customers, you've just given a 10% margin haircut on coffee they'd have bought anyway.

The economics only work if the program acquires new customers, or accelerates the visit rate of customers who'd come less often without it.

What actually drives retention

Three retention drivers, in order of impact:

1. Recognition. The barista who remembers a customer's name and order on the second visit drives more repeat business than any loyalty program. This is unmeasurable but real. A loyalty card adds 0% on top of this; nothing replaces it.

2. Consistency. Customers return when the coffee is the same every visit. Quality variability — a great cortado on Monday and a bitter one on Wednesday — destroys loyalty faster than any program rebuilds it.

3. Pricing fairness. Customers tolerate higher prices when they're consistent and explained. A loyalty discount on inconsistent pricing reads as a guilt offering.

If any of these three is weak, fix them before adding a loyalty program. A loyalty program is a multiplier on retention, not a substitute for it.

Punch cards — when they make sense

Physical punch cards have one strong virtue: they cost almost nothing to run. Print 500 for $40 at the local print shop, stamp them by hand. No software, no fees, no data.

The trade-off: zero data. You don't know who your loyalty customers are, what they order, or how often they come. Without data, you can't measure whether the program is driving incremental visits or just discounting your existing regulars.

Punch cards work best for:

Small cafés where the operator knows every regular and doesn't need data.
Locations with a transient customer base (tourists, conference traffic) where digital signup friction kills participation.
Cafés with a strong "neighbourhood spot" brand where the physical card is a community signifier.

App-based loyalty programs

The alternative is a digital loyalty platform — Square Loyalty, Toast Loyalty, Fivestars, Stamp Me, or a custom built into your POS. These add per-customer cost but generate data.

Typical costs:

POS-integrated (Square, Toast): $25-50/month plus per-customer or per-redemption fees.
Third-party (Fivestars, Stamp Me): $100-300/month based on customer volume.
Custom (built into your own app or POS extension): one-time development, then minimal ongoing.

The value is the data: which customers visit weekly vs. monthly, what they order, when they stop coming. That data, used well, lets you market specifically to lapsed customers ("we miss you — here's a free drink on us if you come this week") which is far more effective than discounting everyone.

The mechanic that works best

The most effective loyalty mechanic we've seen for specialty coffee is the tiered points system with a curated reward tier:

1 point per dollar spent.
50 points = a free standard drink (cortado, latte).
150 points = a free 250g retail bag.
500 points = a private cupping for 6 (or similar high-perceived-value, low-marginal-cost reward).

The bottom tier (free drink) feels accessible. The middle tier (retail bag) drives retail attach for customers who might not have tried your beans at home. The top tier (private cupping) creates an aspirational reward that drives long-term engagement — and the marginal cost to your café is minimal, since the event drives community and word-of-mouth.

The mistakes that make loyalty programs lose money

Too easy. "Buy 5, get 1 free" gives away 20% margin. Avoid.
Discount-shaped. "10% off every visit" trains customers to expect lower prices and damages your pricing position. Worse than no program.
Hard to redeem. If customers have to remember to bring the card, ask for the reward, and navigate awkward conversations to redeem, participation drops to nothing.
No expiration. Points that never expire accumulate liability on your books. 12-month expiration is industry standard and reasonable.
Bolted on without staff training. Baristas who don't know how the program works, or who forget to scan customers in, kill the program. Train, then audit.

The community-loyalty alternative

An interesting third option: skip the points entirely and run a community-loyalty model instead. Examples:

"Coffee passport" for regulars: a card customers fill out by trying each single-origin you serve over a season. Completing the passport earns a free bag or an event invite. Drives experimentation and brand engagement, not just visit frequency.
Subscription: $40/month for unlimited filter coffee, or $60/month for unlimited espresso. Locks in customer revenue, drives daily visits, and creates a sense of belonging. Works best for cafés with a strong daily-regular customer base.
Membership tier: annual fee ($100-$200) for perks — priority on event tickets, first access to new releases, monthly retail bag discount. Generates upfront cash and selects for high-LTV customers.

These alternatives don't fit every café, but they sidestep the discount-shaped trap and align loyalty with relationship rather than transaction.

How to evaluate after 6 months

After running any loyalty program for six months, calculate two numbers:

Cost of loyalty: total value of rewards redeemed.
Incremental revenue: revenue from loyalty members above their projected non-loyalty baseline.

The incremental revenue is the hard part — you have to estimate what loyalty members would have spent without the program. A reasonable approximation: compare their 6-month visit frequency to comparable non-loyalty customers in the same period.

If incremental revenue is more than 3× the cost of loyalty, the program is working. Below that, reconsider.

The simplest advice

If you're a new café, skip a formal loyalty program for the first 6-12 months. Spend that energy on recognition, consistency, and pricing. Once those three are solid, add a simple program and measure honestly.

For more on the customer-economics decisions that intersect with loyalty, see menu pricing and profitability strategies.