The economics of the returning customer

Every pitch for a loyalty programme opens with the same statistic, and almost nobody quoting it has read the paper it came from. We went and read them. What follows uses only figures we could trace to the study that produced them, with the link attached — and it turns out the famous numbers are the weak ones, while the genuinely useful evidence is sitting in two field experiments nobody mentions.

The most-quoted number in loyalty, correctly attributed

The claim is that a 5% increase in customer retention raises profits by 25% to 95%. It is real, it is quotable, and it is almost always credited to the wrong paper.

The 25–95% range comes from Frederick Reichheld and Phil Schefter’s “E-Loyalty: Your Secret Weapon on the Web”, in the July–August 2000 Harvard Business Review. The sentence reads: “By retaining a mere 5% more customers, e-companies can boost profits by 25–95%.” E-companies. In 2000. It is a claim about the dot-com economy at the height of the dot-com economy.

The paper usually credited instead — Reichheld and Sasser’s “Zero defections: quality comes to services”, HBR 1990 — says something narrower and more interesting:

85%

A bank, an insurance brokerage and a car-servicing chain. Every business in that lineage has contracts, switching costs and annual renewals. A café has none of the three — a customer defects by walking past. The direction of the finding almost certainly holds. The magnitude was measured somewhere else entirely, and anyone quoting it at you about coffee is quoting an analogy.

The companion claim has no source at all

The other line you will hear is that acquiring a customer costs five times more than keeping one. We tried to find the study. So, apparently, did Harvard Business Review, which in 2014 printed this:

“Depending on which study you believe, and what industry you’re in, acquiring a new customer is anywhere from five to 25 times more expensive than retaining an existing one.”

Read the construction. “Depending on which study you believe” — and then no study is named, in HBR, in an article whose entire subject is that number. A range that runs from five to twenty-five is not a measurement, it is a shrug. We could not trace it to a primary source, so it does not appear anywhere in this article as a fact.

And there is honest evidence on the other side

Werner Reinartz and V. Kumar studied more than 16,000 customers across four companies over four years and published the result in HBR in 2002. Their finding, verbatim: “we discovered little or no evidence to suggest that customers who purchase steadily from a company over time are necessarily cheaper to serve, less price sensitive, or particularly effective at bringing in new business.”

That is worth sitting with. Loyalty is not automatically profit, and a regular is not automatically a good customer. Which is exactly why the rest of this piece is about mechanisms that were measured rather than slogans that were repeated.

What was actually measured, in a café

In 2006 Ran Kivetz, Oleg Urminsky and Yuhuang Zheng published a study in the Journal of Marketing Research built on a café loyalty programme at a large East Coast university. Buy ten coffees, get one free — the ordinary card. They collected 949 completed cards recording roughly 10,000 coffee purchases.

20%

Customers sped up as the free coffee got closer. Across the whole card the effect is worth about five days: the average card was completed in 24.6 days, against the 29.4 days it would have taken at the pace of the very first gap — a 16% compression.

This is the finding that matters most for a small shop, because it says something the folklore does not. A loyalty card is not merely a record of visits that were going to happen anyway. Within the same set of customers, the card changed the spacing of the visits. And the mechanism — a goal getting nearer — only works if the customer can see how near it is.

Progress the customer cannot see is progress that cannot pull them back in. That is not a marketing opinion; it is what the acceleration is made of.

The two free stamps

The companion experiment is even blunter. Joseph Nunes and Xavier Drèze ran a field study at a professional car wash in 2004, published in the Journal of Consumer Research. They handed out 300 loyalty cards in two versions. One required eight purchases and arrived empty. The other required ten but arrived with two stamps already on it. Both cards demanded exactly eight more washes. The only difference was the framing.

34% vs 19%

Nearly twice the completion rate for the same eight washes. A programme that starts at zero starts cold; a programme that starts as already-begun does not.

The most dangerous moment is the reward

The café study also tracked what happened after the free coffee was handed over, and the answer is: everything slowed down. The last two gaps between purchases on the first card averaged 2.2 and 2.1 days. The first two gaps on the second card were 3.1 and 2.7 days — back to where the customer started, before accelerating again.

The authors call it postreward resetting. In a shop it looks like this: you hand over the free coffee, everyone is pleased, and the customer’s next visit is a day and a half later than the one before. The moment your programme has most earned a customer’s goodwill is the moment it is quietly least able to hold them, and almost no small programme does anything at all on that visit.

Who a programme actually moves

The last piece of evidence is the least flattering and the most useful. Yuping Liu tracked two years of a convenience-store loyalty programme — 42,788 purchases — for the Journal of Marketing. From the abstract: consumers who were heavy buyers at the start “were most likely to claim their qualified rewards, but the program did not prompt them to change their purchase behavior. In contrast, consumers whose initial patronage levels were low or moderate gradually purchased more and became more loyal to the firm.”

2.59 → 4.42

So the returns come from the middle of your customer base and the costs come from the top. Your best regulars claim the most rewards and change their behaviour the least — you are, to a degree, paying them for what they were doing anyway.

That is not an argument against running a programme. It is an argument for knowing which customers are which, which is the one thing a paper card can never tell you.

And locally

One Serbian figure, because it frames the size of the prize. The Statistical Office of Serbia’s Household Budget Survey put average household consumption at 98,165 dinars a month in 2024, of which restaurants and hotels accounted for 3,199 dinars.

3.3%

That is not a share anyone wins with a discount, because a discount is trivially matched by the café across the street. It is won by being the default — the place someone walks into without deciding to. Which is, in one sentence, what all the research above is describing.

What this means for your café

Four practical conclusions follow directly from the studies above, and Coteria is built around them:

• Make the progress visible. The acceleration effect depends on the customer seeing how close they are. On a Coteria pass the stamp count or points bar is drawn on the card itself, in the wallet, so they see it without asking.
• Do not start at zero if you can help it. The car-wash result cost nothing to produce. A welcome stamp on joining is a setting, not a project.
• Do something on the visit after the reward. That is the gap the café study found, and it is precisely the sort of thing an automated nudge exists for. Wallet push campaigns are on the Pro plan.
• Know who is in the middle. The programme’s return comes from occasional and moderate customers, and its cost from your heaviest. The analytics dashboard shows visits by week and member activity — that distinction is the whole reason to have numbers at all.

None of this makes a loyalty programme a growth strategy on its own. The research is fairly clear that programmes are defensive instruments: they change how often the people who already know you come back. That happens to be exactly the problem most cafés actually have.