A feature report says the new product comparison tool received 21,000 clicks in a month. The tool was available to 40,000 users, so someone writes down a 52.5% click rate. The number looks like strong adoption, but it divides events by users and says nothing about whether anyone compared anything.
Adoption needs three separate measurements: who reached the feature, who used it for its purpose, and whether the result changed. This article defines each one, then follows a hypothetical funnel from the click count to a much smaller and more useful answer.
Define the feature's job before counting
Finish this line: "This feature exists to help a user..." For a comparison tool, the ending might be "choose between similar products".
That sentence gives you three definitions:
- Eligible user: someone who could have used the feature in the period. A comparison tool cannot help a user who never saw a category with two comparable products.
- Use: the first point where the feature has done its job for the user. This is usually a system state, such as a table displayed or a result returned.
- Adoption rate: users who reached that state, divided by eligible users, for one stated period.
Both parts of the rate should count users.
Separate the visit, the use and the outcome
Here, a visit means reaching the feature. It does not mean a session. A use means the feature did its work. An outcome means the thing the feature was built to improve happened afterwards.
