, Retail
A coffee shop in Leicester gives out a stamp card: buy nine, get the tenth free. Within four months about 40% of its transactions involve a card. The owner is pleased, because that looks like loyalty.
What has actually happened is that the people who came in four times a week before are still coming in four times a week, and now every tenth cup is free. He has given a 10% discount to his most committed customers and changed nobody's behaviour. His margin has moved and his footfall has not.
A scheme that rewards what you already had is a price cut
Start with the only question that matters: what behaviour do you want that you are not getting.
The answers are usually specific. You want the customer who visits monthly to visit fortnightly. You want the customer who buys one category to try a second. You want the Tuesday to look more like the Thursday. You want the basket to go from £18 to £24. You want to know who your customers are so you can tell them something in November.
Each of those implies a different scheme, and none of them implies a flat percentage back on everything. A reward tied to frequency, where the benefit resets if you do not come in within a period, changes frequency. A reward that unlocks when a customer buys from two departments changes range penetration. A threshold reward at a basket value slightly above your current average changes basket size. A flat points scheme changes nothing except your gross margin.
The cost side is simple to model and rarely modelled. If you give back 5% in rewards and 70% of the spend it applies to would have happened anyway, you have spent 5% of your entire loyal revenue to influence 30% of it. The scheme has to generate enough incremental margin from that 30% to cover the giveaway on the other 70%. That is a high bar, and it is why targeted rewards beat universal ones.
Breakage is a cost forecast, not a windfall
Breakage is the proportion of rewards that are never redeemed. Stamp cards get lost. Points expire. Customers move away.
Every loyalty scheme relies on it to some degree, and it is where the arithmetic goes wrong in both directions. Assume high breakage and you can afford a generous headline reward. Then improve your app, send reminders, and watch redemption climb towards the level you never budgeted for. A scheme costed on 40% breakage that achieves 10% breakage is an unplanned margin cut across your best customers.
The opposite failure is worse commercially. A scheme where rewards are hard to reach and quietly expire produces a customer who has spent a year collecting something and got nothing, which is a customer with a grievance rather than a habit.
Where you accrue a liability for rewards that will be redeemed, it belongs in your accounts. Under FRS 102 you are not recognising the full sale as revenue if part of it is effectively a promise of future goods, and a material reward programme needs the deferral treated properly rather than as a marketing expense. The VAT treatment of loyalty rewards has been litigated repeatedly. Where the reward is a discount you give on your own goods, VAT follows the reduced amount you actually take. Where a third party operates the scheme and you pay them for redemptions, the analysis is more involved and worth an hour of your accountant's time before launch rather than a year after.
The data is usually worth more than the discount
The reason supermarkets run loyalty schemes is not the loyalty. It is that a card links every basket to a person, and that turns a pile of transactions into a customer history.
For a small retailer the same principle applies at a smaller scale and is more actionable, because you can act on it personally. Knowing that 300 of your customers bought a particular category last spring and only 90 of them have bought it this year is a specific commercial fact you can do something about within a week. A scheme that produces that is earning its keep even if the rewards themselves do nothing.
That comes with obligations. Under UK GDPR you need a lawful basis for processing, you have to tell people what you are collecting and why in a privacy notice they can actually find, and you need to be able to delete it on request. If you are processing personal data you will generally need to pay the data protection fee to the Information Commissioner's Office.
Marketing by email or text is governed separately by the Privacy and Electronic Communications Regulations. For consumers the default is consent. The soft opt-in exception lets you email existing customers about similar products where you collected the address in the course of a sale and gave them a simple chance to refuse, in that message and every subsequent one. A tick box on a loyalty form that is pre-ticked is not consent, and bundling marketing consent into the terms of joining the scheme is not consent either.
Keep the data you will use. A scheme that collects date of birth, address and phone number and then sends two emails a year has taken on risk for nothing.
Judge it on cost per retained customer, not on sign-ups
Sign-ups are the vanity number. Every scheme signs people up, because the till operator asks and people say yes.
The numbers that tell you whether it works are these. Active rate: the proportion of members who have transacted in the last ninety days. Repeat frequency for members against non-members, measured before and after they joined rather than against each other, because members were always your better customers. Basket value for the same. Redemption rate. And total reward cost divided by the number of customers who moved from occasional to regular.
Run it for a year before deciding, because retail is seasonal and a scheme launched in October always looks good by December.
Then be willing to kill it. A scheme that is not changing behaviour is a permanent discount with administration attached, and closing one is awkward. Give notice, honour outstanding rewards, and do it in a quiet month.
The alternative worth considering first costs nothing. Recognising a regular customer, knowing what they buy, and holding something back for them is loyalty that no app replicates, and independent shops are the only ones who can do it.
Where this shows up in the game
Shops are part of Business Baron's full game, and e-commerce is free to run. Marketing is campaigns with a target and a budget, and attribution misleads you the way real attribution does, which is exactly the problem with judging a loyalty scheme by who uses it. A reward comes out of margin immediately and comes back slowly, if at all.