The Stickiness Test
Judge a new product by frequency of use and reversion, not launch sales
- Difficulty
- Easy
- Time to result
- ~weeks to results
- Steps
- 4
- Confidence
- 84%
The Stickiness Test says the questions that predict a new product's fate are how frequently it is used and, of the people who try it, how many revert. Certain products cannot be judged in advance because owning or using them changes your utility function: an air fryer, a Japanese toilet seat, multi-channel TV, an electric car. Nobody wants them until they try one, then almost nobody goes back. High-frequency products, used daily or every two days, reset preferences fastest, which is why frequency matters so much. Because behaviour change follows a slow-starting sigmoid curve driven by social copying and habit, early flatness fools people into abandoning products too soon. So stop asking 'how fast are we growing?' and ask 'how sticky is this, and how many adopters ever revert?'
Origin
Rory Sutherland's frequency-and-reversion test for new products, drawn from Dragons' Den's Hungry House miss and examples like air fryers and electric cars. Extracted from Deep Dive with Ali Abdaal.
Core principles
- 01Some products cannot be judged in advance because using them changes your preferences
- 02Frequency of use predicts adoption better than initial appeal
- 03The real question is not how fast you are growing but how sticky the product is
- 04Behaviour change follows a slow-starting sigmoid, so early flatness misleads
How to run it
- 1
Ask the frequency question
Determine how often the product will genuinely be used: daily, weekly, monthly or yearly.
Pro tip Daily-use products reset preferences fastest and are the best bets.
Watch out A great idea used once a year rarely builds the habit needed to take off.
- 2
Ask the trial question
Find out whether people who use it once go on using it.
Pro tip A strong once-then-repeat signal beats a large but shallow trial base.
- 3
Ask the reversion question
Of the people who adopt it, measure how many revert to the previous alternative.
Pro tip Near-zero reversion means preferences have been permanently reset.
- 4
Reframe growth as stickiness
Replace 'how fast are we growing?' with 'how sticky is this?' and give sigmoid-shaped adoption time.
Pro tip Advertising's job here is to generate trial and accelerate an inherently slow start.
Watch out Assuming adoption is linear makes you give up too soon.
In the wild
Dragons rejected Hungry House, later sold for around £220m, because they argued about whether people would order takeaway online at all. The only question that mattered, Sutherland says, was whether people who used it once, or three times, kept using it. Frequency and retention, not the debate about initial appeal, decided its fate.
→ A frequency-and-retention lens would have flipped a famous investment miss.
Nobody wants an air fryer or a Japanese toilet seat until they try one; then about half become evangelists and almost nobody reverts. These daily-use products reset the user's utility function so completely that going back feels unthinkable.
→ High-frequency trial converts sceptics into permanent, evangelising users.
Common mistakes
Judging by launch sales
Aggregate early sales say nothing about retention; the longitudinal reversion rate is what predicts where the market goes.
Assuming linear adoption
Because significant new products follow a slow sigmoid, treating early flatness as failure kills products prematurely.
Is it for you?
Best for
Founders and product teams evaluating whether a genuinely new product or habit will take off.
Not ideal for
Low-frequency, one-off purchases where trial cannot build a habit or reset preferences.
From the transcript
“The question you need to ask is not how fast are we growing it's how sticky is this product”
“the real question you need to ask if you want to predict where the market's going is of the people who buy electric cars how…”
From the episode
How To Influence People: Marketing Secrets Behind The World’s Biggest Brands - Rory Sutherland
Rory Sutherland