Most businesses either overthink this and set a threshold nobody reaches, or never cost it at all and are surprised later. The calculation takes two minutes and it is the difference between a programme that pays for itself and one that quietly leaks margin.
How do you cost a loyalty reward?
Use the cost of goods, not the menu price, and measure it against what the customer spent at full price to earn it. The menu price is what the reward is worth to the customer; the cost of goods is what it costs you, and the gap between the two is what makes rewards efficient.
| Business | Reward | Menu price | Your cost | Earned on | Real cost |
|---|---|---|---|---|---|
| Café | Free coffee, 10th visit | €4.00 | ~€0.90 | €36 | 2.5% |
| Bakery | Free pastry, 8th visit | €3.50 | ~€0.80 | €24 | 3.3% |
| Sandwich shop | Free sandwich, 8th | €7.00 | ~€2.20 | €49 | 4.5% |
| Restaurant | €10 off at 200 points | €10.00 | ~€3.50 | €200 | 1.8% |
| Salon | Free blow-dry, 6th visit | €25.00 | ~€0 + 20 min | €240 | chair time |
| Pet shop | €10 off at 200 points | €10.00 | €10.00 | €200 | 5.0% |
Real cost of the reward as a share of the spend needed to earn it, from the table above.
Two rows deserve attention. The salon's reward costs almost nothing in materials but consumes a bookable slot, which in a chair-time business is the real constraint. The pet shop's money-off reward costs its full face value, because a discount is not a product, which is why retail rewards should be set as a smaller percentage than hospitality ones.
What percentage is right?
Between 2% and 7% of what the customer spends. Below 2% the reward is too small to change behaviour; above 7% you are funding a discount rather than a loyalty programme, and the customers who cost you most are the ones who were already loyal.
Hospitality can sit at the higher end because cost of goods is low relative to price. Retail that gives money off should sit at the lower end, because every euro of reward is a euro of margin.
Which rewards look generous and cost little?
Anything where perceived value is set by a menu price and your cost is materials or minutes. These are the rewards worth reaching for, because the customer values them at retail while you pay wholesale.
- A drink or pastry: perceived at menu price, costs a fraction of it.
- An add-on service, a nail trim, a de-shedding, a blow-dry, costs minutes rather than a bookable slot.
- Early access to a sale, which costs nothing at all and is valued.
- An upgrade on something already being bought: a larger size, a better cut, an extra topping.
- Meeting-room credit or an off-peak day, where the capacity is going unsold anyway.
The common property is that none of them gives away a unit of constrained capacity. A free haircut costs a salon an hour it could have sold; a free blow-dry costs twenty minutes it probably could not.
Should the reward be the customer's usual order?
Yes, wherever you can afford it. A free filter coffee given to someone who always buys a flat white reads as a downgrade and teaches customers the programme is mean, which costs more in goodwill than the price difference costs in margin.
The difference is usually small once properly costed: a euro at most on a hot drink. Where it is large, the workable middle is a free standard item with the customer paying the difference for an upgrade.
How does the threshold change the cost?
The threshold sets the denominator, so doubling it halves the cost as a percentage, and also halves the number of people who reach it. A reward nobody reaches costs nothing and achieves nothing.
| Threshold | Reward cost as % of spend | Typical completion |
|---|---|---|
| 5 visits | ~5% | High |
| 8 visits | ~3% | High |
| 10 visits | ~2.5% | Good |
| 15 visits | ~1.7% | Falling |
| 20 visits | ~1.2% | Poor: most never finish |
The sweet spot for a frequent-purchase business is eight to ten, because it is reachable in about a month for a regular and still represents real repeat business. Stretching to twenty looks cheaper on a spreadsheet and is usually cheaper because the programme stopped working.
What costs should you include beyond the reward?
The platform subscription, the enrolment messages, and staff time. For most small businesses the subscription dominates and the rest is close to noise.
Wallet loyalty platforms generally run €20 to 90 a month, enrolment texts cost between one and ten cents each depending on country, and the staff time is three seconds per enrolment. Against a reward cost of a few hundred euros a year, the honest total is still a small fraction of the revenue it protects.
Should the reward cost scale with what the customer spent?
With points, yes automatically; with stamps, no. That is the whole difference between the two mechanics, and trying to make a stamp card scale is what produces the unreadable schemes customers stop tracking.
If your tickets vary by more than roughly double, the stamp card is already mispricing the reward: overpaying small baskets and underpaying large ones. The fix is the mechanic rather than the reward value.
What happens to the cost as the programme grows?
Reward cost scales with enrolled customers while the platform cost steps up in bands, so the cost per customer falls as the programme grows and the reward becomes the dominant line. That is the right shape: the variable cost only occurs when a customer has already spent.
| Enrolled | Plan band | Rewards/yr at 2 cards each | Total/yr |
|---|---|---|---|
| 100 | €24.99/mo | ~€200 | ~€500 |
| 300 | €44.99/mo | ~€600 | ~€1,140 |
| 800 | €84.99/mo | ~€1,600 | ~€2,620 |
Those reward figures assume a €1 cost of goods and two completed cards per customer a year; substitute your own. The point of the table is the shape rather than the numbers: the subscription matters at 100 customers and is close to irrelevant at 800.
What is the cheapest reward that still works?
An upgrade on something the customer is already buying: a larger size, an extra shot, a better cut of the same order. It costs the marginal difference rather than the whole item, and it is received as a favour rather than as a transaction.
The limit is that an upgrade is less legible than a free item when you are explaining the programme at a counter. "Ten coffees, the eleventh is free" survives being repeated by a part-time staff member; "a free upgrade after ten" invites a question every time.