What a loyalty programme is really worth

You will have read it somewhere: “a 5% rise in customer retention lifts profits by 25 to 95%”. The source quoted most often is a Fred Reichheld brief published by Bain & Company in October 2001. Open it and this is what it says: in financial services, a 5% increase in customer retention produces more than a 25% increase in profit. Financial services, in 2001. The example from that sector in the document is Vanguard, an American fund manager. The top of the range is not in it at all.

The figure is not wrong. It is simply not about you. A saver leaves money with a fund manager for years without a second thought. Your customer weighs up your window against the one across the road every morning. Your margins, how often people walk in and what they spend have nothing in common with a savings contract.

So you will find no promised result in the rest of this article. Only a sum you can do in ten minutes, on the back of a receipt, with numbers you already know.

A loyalty card brings nobody new through your door. That is the job of your shopfront, word of mouth and where you stand on the street. What a card can change is the rhythm of the people who already know you: the customer who came in twice a week comes three times, and the one hesitating between you and the shop opposite picks your door because they have seven stamps out of ten.

The unit of measurement for a loyalty programme is therefore the extra visit. Not the “loyal customer”, which means nothing and cannot be counted. And a visit is not worth its ticket: it is worth the margin you keep on that ticket. A £6.50 sale with £2.30 of stock behind it leaves you £4.20. That is the number that counts.

The question then becomes very simple, and above all checkable: what is one visit worth here, and how many extra visits can this programme trigger in a month?

You need four of them. The first three are in your till. Your average sale: a normal week's takings divided by the number of transactions. Your margin on that sale: what is left once the stock is paid for. And the number of customers signed up to the programme, or the number you are aiming at for the first three months.

The fourth is the only one you estimate: the share of those sign-ups who will make one extra visit a month because of the card. Be deliberately pessimistic. Start at one customer in ten. If the sum already works at that level, it will keep working. If it only works by assuming that half your customers change their habits, that is not a sum, it is a hope.

The formula fits on one line: the number of sign-ups, times the share who come back one extra time, times the margin on a visit. An example, and these are example numbers, not trade averages: 200 sign-ups, one in ten coming back once more, so 20 visits, at £4.20 of margin, makes £84 a month. Put your own in: with a £40 sale in a hair salon, the same reasoning gives a completely different answer.

This is the line merchants overestimate most. A free coffee at £3 does not cost you £3. It costs you the beans, the milk and the cup. There is one case where you do lose the whole margin: when the customer would have come back anyway. For the one who comes back because of the card, you only pay the cost price of something you would not have sold.

Count the volume, not the anecdote. On a ten visit card, a weekly regular claims a reward every ten weeks. If 40 of your sign-ups come in once a week, that is about four rewards a week, so seventeen in a month. At 55p of cost price each, rewards weigh around £9.50 a month. But those seventeen rewards do not all go to customers who came in on top: for the ones who would have come anyway, it is the whole margin on the free product that walks out, £2.45 on a £3 coffee instead of 55p. Count them at their worst, full margin for everybody, and the £84 of the previous example drops to just over £40. If the sum still stands there, it will stand for real.

The common sense rule that follows: if the cost of the rewards is greater than the margin on the extra visits, your threshold is too low or your gift too generous. Raise the threshold by a visit or two, or change the reward, then do the sum again. It is also why a free product beats money off: it costs you less and it pleases more.

On a free plan there is no subscription to earn back. Digikart Free is €0, with no bank card and no cap on customers: your only outgoings are the rewards and the few seconds of scanning at the till. The bar to clear comes down to the cost of the rewards you hand out.

For a paid plan, turn the sum around instead of asking yourself whether “it works”. Divide the subscription by your margin per visit, before rewards: you get the floor, the number of visits below which the subscription is certainly not paying. Digikart is billed in euros, in the United Kingdom as everywhere else. With £4.20 of margin, the Plus plan at €17 a month incl. VAT, around £15 at September 2026 rates, is paid back by four extra visits in the month; the Pro plan by eight at its launch price of €39 incl. VAT, in force until 31 December 2026, and by ten at its standard €49. In Ireland there is nothing to convert: €17 against €4.20 of margin gives the same four visits. Over a year, €170 for Plus and €390 for the first year of Pro come out at the same orders of magnitude.

Put that way, the question becomes concrete: can notifications, a flash deal on a dead Tuesday and a win-back of the customers who have not been in for six weeks bring me ten visits in a month? You know your regulars, you can answer that. And if you are not sure, that is what the trials are for: three days on the Plus plan, fourteen days on the Pro plan, with no commitment.

A card already started gets finished more often than a blank one. That is not shopkeeper's intuition: Joseph Nunes and Xavier Drèze measured it in the Journal of Consumer Research in March 2006. Turning an eight visit card into a ten visit card with two visits given away up front raises the odds that the customer goes all the way, and shortens the time they take over it. The effort asked for is exactly the same. The welcome bonus on the Plus plan is there to set up precisely that: you raise your threshold by two visits and you give those two visits away at sign-up.

The second lever is a reachable threshold. A regular has to get their reward within one to two months, otherwise they forget the card before the end. When the cycle is long, tiers in between stop the gap opening up: up to three tiers in the same cycle on the Plus plan, with the customer choosing which tier to stop at, so there is always something close in front of them.

The third lever is the one that pays the most into the sum, because it goes looking for visits that would not have happened: the win-back of inactive customers, the flash deal on a quiet hour, the birthday reward. On the Pro plan these notifications are unlimited and carry no billed SMS, so they add nothing on the spending side. They only move the number of visits.

It will not tell you why a customer came back. Nobody can untangle the loyalty card, the weather, the price across the road and a sudden want of a sausage roll. Your sum gives you an order of magnitude and a decision threshold, not proof. Be wary of anyone promising you a percentage of extra takings: they know neither your margin nor your customers.

Nor will it tell you what share of your rewards goes to customers who would have come back with nothing at all. That share exists, it is normal, and it is why the cautious version counts every reward at full margin: you cannot know which ones were earned. What you look at in the end is the balance, not perfect targeting.

To settle it for good, take three figures after two months: the number of sign-ups, the share of sign-ups who have come back at least once since signing up, and the number of rewards claimed. The dashboard shows sign-ups and rewards; the share of customers who have come back at least a second time is the retention rate in the detailed statistics, on the Pro plan, with the CSV export to compare one month against the next. If the second visit does not take off, the problem is almost always the threshold or the sentence said at the till, and rarely the tool.

Then comes the decision, and it is yours. Do this sum with your four numbers. If it comes out derisory, launch nothing, you will have saved yourself the time. If it stands up even on cautious assumptions, start with the free plan: you will swap your estimates for your real figures without spending a thing.