Toothbrush Test
Solve a major recurring problem so well that people return every day
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 4
- Confidence
- 98%
The Toothbrush Test is Larry Page's product-selection rule as recounted by Gawdat: find a major problem, solve it well, and favor something people use as routinely as a toothbrush. The test combines problem magnitude with recurring utility. A serious problem creates value, repeated use keeps the solution close to the user's life, and excellent execution can create a business without merely cloning a crowded competitor. Apply it by defining the problem, verifying how often it occurs, identifying the failure in current alternatives, and measuring whether the solution earns repeated use because it helps. The toothbrush is an analogy rather than a literal requirement. Some excellent businesses solve infrequent problems, so frequency should be interpreted in relation to the market rather than used as an automatic rejection rule.
Origin
Gawdat attributes the Toothbrush Test to Google co-founder Larry Page, whom he describes as one of the most intelligent people he encountered. Extracted from The Diary of a CEO.
Core principles
- 01A major problem creates room for meaningful value
- 02Frequent use compounds usefulness and habit
- 03Solving well matters more than copying a crowded feature
- 04Strong value creation can reduce dependence on direct rivalry
How to run it
- 1
Choose a major problem
Define a consequential user problem rather than beginning with a fashionable feature or competitor. Specify who experiences it and what the current failure costs them.
Pro tip Use observed behaviour or direct customer evidence to establish the problem.
Watch out A large market does not prove that the specific problem is urgent.
- 2
Test recurrence
Determine how often the user encounters the problem and could benefit from a solution. Frequent recurrence can compound value and create a natural reason to return.
Pro tip Measure the natural rhythm of the problem rather than manufacturing engagement.
Watch out Do not optimize addiction or empty usage simply to raise frequency.
- 3
Solve the core well
Build around the central outcome and remove distractions that do not improve it. Compare the result with the user's current workaround.
Pro tip A narrow solution to a painful recurring problem can outperform a broad imitation.
Watch out Novelty without a better outcome does not pass the test.
- 4
Measure earned return
Observe whether people come back because the product repeatedly solves the problem. Pair usage with evidence of practical benefit.
Pro tip Ask returning users what would become harder if the product disappeared.
Watch out Retention alone can reflect lock-in rather than genuine value.
In the wild
Gawdat recalls Page asking why someone would compete on another photo-sharing app when they could find a major problem and solve it so well that people use the solution twice a day, like a toothbrush. The example is a product-selection analogy rather than a promise of commercial success.
→ The builder prioritizes a consequential recurring need over copying a crowded product category.
Common mistakes
Starting with the competitor
Copying an existing category can obscure whether a meaningful user problem remains unsolved.
Manufacturing frequency
Notifications and compulsive loops can increase use without increasing value. The recurrence should come from the problem.
Treating frequency as universal
Some high-value problems occur rarely. Adapt the test rather than rejecting them solely for low usage frequency.
Is it for you?
Best for
It is best for evaluating startup and product ideas before committing significant build time or capital.
Not ideal for
It is not ideal for rare but valuable purchases where daily use is irrelevant to customer value.
From the transcript
“Larry used to call it the toothbrush test.”
“find a major problem and solve it really well”
“people use you quite a you know, twice a day”
From the episode
EMERGENCY EPISODE: Ex-Google Officer Finally Speaks Out On The Dangers Of AI! - Mo Gawdat