Sample frameDay 201·Marketing · Locate

Morning. Your repeat rate isn't a verdict on your product. It's a verdict on your size.

Marketing.

Double Jeopardy

5 min read·Apply by lunch

The question

Pull up your repeat-purchase rate, your churn and your average orders per customer — and before you judge a single one of them, can you say what those same numbers look like for the brand one rung above you in share, and the one just below? Without that comparison you are reading your own size and calling it your product's character.

The idea

Double jeopardy was named by William McPhee, a sociologist at Columbia, in Formal Theories of Mass Behavior (1963). He noticed that less popular radio programmes and comic strips were hit twice over: fewer people knew them, and the people who did know them liked them slightly less. Andrew Ehrenberg, the British statistician who spent a career reading buying data at London Business School and later South Bank University, kept finding the same shape in brands. With Gerald Goodhardt and Patrick Barwise he set it out in the Journal of Marketing in 1990, under the title Double Jeopardy Revisited: across categories, countries and decades, a brand with a smaller share of its market has fewer buyers, and those buyers buy it slightly less often. Note the order of magnitude, because the whole diagnostic follows from it — the gap in buyers is wide, the gap in frequency is narrow. The second penalty is the one that gets misread. A brand at four per cent share looks at its repeat rate, concludes it has a loyalty problem, and spends the year building a rewards programme, when that repeat rate is precisely what a four-per-cent brand should have. Loyalty follows share. It doesn't lead it.

The framework

01
Build the ladder before you read your own number.
Take the five or six brands in your category and set share alongside whatever loyalty measure you actually hold — repeat rate, purchases a year, share of requirements. The ladder appears immediately and your own number stops floating free. Most teams have only ever compared themselves with the leader, which guarantees the wrong verdict: the leader's loyalty is high for the same reason its share is. What you need is the rung above you and the one below, and an hour with whatever category data you can beg is enough to draw them.
02
The wide gap and the narrow one.
Read the ladder down both columns and notice how differently they behave. Share and buyer counts fall away steeply as you descend; purchase frequency moves far less, so the spread from the largest brand to the smallest is a fraction of the spread in buyers. That asymmetry is the entire finding, and it is what turns the ladder into a diagnostic rather than a curiosity. If your frequency sits close to everyone else's while your buyer count is a tenth of theirs, the product isn't the thing that's wrong. What to do about a buyer gap that wide is Day 43's argument, not this morning's.
03
A normal repeat rate is not a compliment.
The verdict has two readings and both are useful. Sit roughly where a brand your size should sit and your loyalty is normal — neither praise nor a problem, but it does mean the product is fine and too few people think of you or can easily buy you. Sit clearly below the line and you have found the rare, genuine retention fault, worth a quarter of real work. The failure is reading a middling repeat rate as evidence of a middling product, then spending a year repairing something that isn't broken.
04
Two exceptions, and how to count the second one.
Double jeopardy describes how markets behave; it isn't permission to neglect customers, and the exceptions are specific. One is the case above — sitting clearly below the line for your size, which is a genuine retention fault and the reading that earns a real retention quarter. The other is subscription, where a cancellation removes a buyer outright rather than lowering their frequency, so defection is a loss of buyers wearing other clothes. That one has to be booked in the buyer column of the ladder rather than the frequency column. Put it in the wrong column and the ladder hands you a frequency verdict on what is plainly a buyer number.

Members only · 4 principles + template + AI mentor

364 more frameworks are waiting.

Double Jeopardy is Day 1 of 365. One framework every morning for a year — across strategy, sales, negotiation, leadership, and more.

$1 /day

Billed $365/yr · cancel any time

365 frameworks

12 topics · templates · AI mentor

Reserve your place — $365/yr →

One payment a year · the whole library · no auto-renewal.

More Marketing frameworks

All 28 marketing frameworks →