In brief
A shop cannot fund a free item the way a cafe funds a free coffee. At a 40% gross margin every pound you give away at cost looks like £1.67 to the customer, against four pounds in hospitality, and the rewards that survive that are credit, access and the one item on your shelves with the widest gap between price and cost.
The short answer
Retail has less room to fund a reward than any other local trade, which is why “tenth item free” is normal in a cafe and rare in a shop. A cafe gives away eighty pence of coffee that sells for £3.20. A shop giving away a £30 item gives away £18 of stock, and the customer can read the shelf price. The fix is not a longer card: count cumulative spend rather than visits, and make the reward a fixed value or a privilege rather than a product the customer picks.
The pot the reward comes out of
Gross margin per visit, multiplied by the number of visits on the card. That is the whole budget for a retail loyalty program, and it is smaller than it feels.
Take a gift shop with a £25 average basket and a 40% gross margin. Each visit earns £10 of margin. Ten visits earn £100. A free £30 item costs £18 in stock, which is 18% of that pot, and the margin test in how to reward customer loyalty puts the comfortable ceiling nearer 10%.
Getting £18 under 10% needs £180 of margin, which is eighteen visits at £25. For a customer who comes in monthly, that card takes a year and a half. Nobody finishes it, so the program promises a great deal and pays out almost nothing, which is the worst combination on offer.
The words vary by market. Retail rewards programs, retail loyalty schemes, loyalty programmes in retail with the British spelling: the arithmetic is identical in all of them.
Cost and shelf price sit close together in retail
The second number is what makes retail different rather than merely poorer: the retail value of the reward divided by what it cost you.
For a £3.20 coffee that costs £0.80, that multiple is four. Every pound of cost of goods buys four pounds of perceived value. In a shop running a 40% gross margin the multiple is 1.67, and at 30% it is 1.43. It is fixed by your margin, and no reward design changes it.
Hospitality manufactures perceived value out of cheap inputs. A shop that buys finished goods and resells them cannot, which is also why the strongest reward class of all, giving away something you make yourself, is unavailable to most retailers. You do not make anything.
| Shop | Basket | Gross margin | Margin per 10 visits | Reward at 10% of it | What that buys at your own prices |
|---|---|---|---|---|---|
| Greengrocer | £12 | 30% | £36 | £3.60 of stock | About £5 on the shelf |
| Bookshop | £18 | 35% | £63 | £6.30 of stock | About £10 on the shelf |
| Gift shop | £25 | 40% | £100 | £10.00 of stock | About £17 on the shelf |
| Cafe, for contrast | £3.20 | 75% | £24 | £2.40 of stock | About £10 on the menu |
Baskets and margins here are assumptions to be replaced with yours, not findings about the trade. The point is the last column. Ten visits to the greengrocer buy the customer five pounds of fruit. Ten visits to the cafe buy three coffees. Same discipline, and one of those gifts looks generous.
Count spend, not visits
In retail the visit is the wrong unit, because two visits are not worth the same to you. One customer spends £4 and another £140, and a stamp cannot tell them apart.
A minimum qualifying spend patches this, and where that threshold belongs is worked through on loyalty programs for retail stores. Cumulative spend solves it properly: £250 spent, £10 back. That reward is 4% of the revenue and, at a 40% margin, exactly 10% of the £100 of margin behind it. The card is now funded by the thing that actually funds you.
Cumulative spend has a second useful property. One £250 trip and twenty-five £10 trips arrive at the same place, which is honest, because both were worth £100 to you.
The cost is operational. A stamp needs no number; a spend total needs the amount typed in at the till or passed from it, every time, by whoever is serving. That is the real trade-off between the two mechanics, and there is more of it in stamp cards versus points.
Pay in credit, not in stock
A fixed-value credit has a ceiling. A free item does not, because the customer picks the item.
Run the gift shop both ways. Free £30 item: £18 of stock leaves the shop and nothing comes in, so the reward cost £18, and it costs more the moment somebody picks something dearer. Capping it means writing “up to £30 in value”, which is a credit wearing a costume.
Now the £10 credit against a £30 purchase. The customer pays £20, you part with £18 of stock, and you book £2 instead of £12. The reward cost £10 of forgone margin, and that is what it costs whatever they spend it on. A credit costs exactly its face value, never a penny more, as long as you set two rules: no change given, and it is used in a single transaction.
That does contradict something. The advice not to hand out cash equivalents is right about gift cards, and a store credit sits uncomfortably close to one. Two things separate them: a credit can only be spent with you, and it is only ever issued against spending that already happened. It is still the weakest option on this page, and it is here because the strong option is not available to a shop.
Give access, which costs you no stock
Access is the one reward class where retail is not at a disadvantage, because nothing leaves the shelf.
- First refusal on limited stock, or a 24-hour hold on something that just came in. Costs nothing when supply is loose, and it is worth the most when it is tight.
- The sale before the sale. Members buy the markdown a day early. The discount is already funded, since the stock was going to be reduced anyway, and the only real cost is the handful of items that would have sold at full price during that extra day.
- A message when the thing they asked about comes back in, sent to one person rather than to a list.
- Ordering in something you do not stock, without asking for a deposit.
Access costs capacity rather than goods, so it gets more expensive exactly when you are busiest, which is also when the customer values it most. The one thing it needs is somebody remembering, so write down who and when before you promise it.
Find the item with the widest gap
The multiple is a property of an item, not of a shop. Find whatever on your shelves has the biggest gap between price and cost, and make that the reward.
Most shops have two or three: an own-brand line, a consumable, an accessory, the packaging, something a supplier left you for nothing. A candle at £12 that cost you £3 has a multiple of four, which is cafe territory sitting inside a shop that averages 1.67.
One rule about which item to pick. It should never be the thing the customer came in for. That one you were going to sell at full price, and giving it away swaps £12 of margin for goodwill you could have bought for £3.
The card has to be produced before the till closes
Retail loyalty programs fail at recognition more often than at arithmetic. A shop visit is anonymous by default, and the card only counts if the customer produces it in the seconds between “that’s £24.60” and the card machine beeping.
A cafe knows its regulars by face. A shop serving a hundred people a day, most of them strangers, does not, and nothing about the transaction requires anyone to say who they are. Contactless, receipt, gone.
Three decisions make this survivable:
- Write down the forgotten-card rule before it happens. Someone reaches the door and remembers. “We can add it from today’s receipt, not next week’s” is a rule anyone on shift can apply without calling you. An unwritten rule becomes an argument in front of a queue.
- Keep the retrofit off the counter. Adding a missed stamp afterwards is back-office work: in NeoLoyal it is a manual adjustment on the customer’s record in the owner dashboard, deliberately not something the till does mid-queue.
- Do not build it on a phone number lookup if your queue moves quickly. Typing eleven digits takes longer than the rest of the transaction, and staff quietly stop asking.
The join moment matters more here than the reward does, and the sequence for setting one up is in the guide to planning a visit-based program.
Three numbers to take from last month’s till report
Average basket. The gross margin you actually make on it, after the discounting you actually did. The number of visits your best fifty customers made.
Multiply the first two, multiply by the third, then take a tenth. That figure is what a reward can cost you at cost of goods without the program becoming a discount scheme with stamps on it. If it does not buy anything worth having, you have your answer: the reward is a credit or a privilege, and the card was never the problem. The rest of the budget, including the part that grows when the program works, is in what a loyalty program actually costs.