The Repeat-Purchase Meter
Judge product value by whether experienced buyers willingly buy again
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 3
- Confidence
- 95%
The Repeat-Purchase Meter evaluates value after customers have experienced the product. Define an appropriate replacement or renewal window, identify buyers who were eligible to purchase again, and calculate how many chose the same product or at least stayed within the category. Add a counterfactual question: if the item broke tomorrow, would the customer replace it immediately? This distinguishes products whose promise attracts initial buyers from those that become useful parts of life. Sutherland argues that companies overinvest in acquisition because it is fast and easy to quantify, while retention is slower and harder to observe. Marketplaces could help consumers by displaying repeat behavior alongside star ratings, provided they exclude products not naturally repurchased and account for consistency bias or sunk-cost effects.
Origin
Sutherland proposes a repeat-purchase indicator for marketplaces, contrasting an air fryer that owners would quickly replace with appliances that are bought once and then abandoned.
Core principles
- 01Acquisition is easier to count than retention but often less informative
- 02Repeat purchase reveals value after the promise has been tested
- 03The metric must exclude genuinely one-off categories
- 04Long time horizons reward retention investment
How to run it
- 1
Define eligibility
Set the time window and event that makes another purchase plausible, such as renewal, depletion or product failure. Exclude customers without a genuine opportunity to repurchase.
Pro tip Use category-specific windows.
Watch out A universal window will misclassify slow and fast replacement cycles.
- 2
Measure experienced choice
Track the percentage of eligible prior buyers who purchase the product or remain in the category again. Separate repeat behavior from first-time acquisition.
Pro tip Pair behavioral data with the 'replace tomorrow' question.
Watch out Repurchase may partly reflect switching costs rather than satisfaction.
- 3
Rebalance investment
Identify product and service features associated with repeat use, then compare retention investment with acquisition spending. Shift resources toward durable experienced value.
Pro tip Report acquisition and repeat rates side by side.
Watch out Optimizing retention should not conceal a shrinking inflow of suitable new customers.
In the wild
Sutherland suggests asking whether owners would buy another appliance the day after it broke. Many air-fryer users would replace it immediately because it has become useful, while a neglected yogurt maker may never be purchased again.
→ Repurchase intent distinguishes sustained utility from an attractive initial promise.
Common mistakes
Using repeat rate on one-off purchases
Low repurchase means little when the product is designed to last a lifetime or solve a singular need.
Reading all repeat behavior as delight
Consistency bias, sunk costs and switching friction can retain customers even when satisfaction is mediocre.
Is it for you?
Best for
Repeatable consumer categories where customers have enough time and opportunity to choose the product or category again.
Not ideal for
Marriage, houses and other genuinely rare or one-off decisions where repurchase is not a meaningful success signal.
From the transcript
“Nearly all businesses overinvest in acquisition and underinvest in in in um customer retention”
“It's not just how many people bought this thing. It's how many people that bought this thing before rebought.”
From the episode
Waymo, Texas Culture, Airline Lounges, OpenAI & Uber Eats - Rory Sutherland - #973 -