Product-Market Fit describes a product's propensity to retain and sustainably grow in a market niche, relative to competitors. Fit products are like fit organisms; they must survive (aka retain) and reproduce (aka grow) in at least one market niche.

The first thing to focus on is retention, a measure of whether new users stick around. It's defined as the share of users who are still active after a specific time period following their first action date.

To understand what causes retention, you can build a model. Begin by putting your data into the following format:

Create a row for each user, a column for the signup date, a column for whether those users retained, and additional columns for each variable we think could explain why a user might have retained. For instance, we could add a column for phone type if we think our Android build is buggier than iOS.

Next run a logistic regression, including only users who signed up in the same date cohort. In the R example below, we are continuing on with our phone OS example and two more explanatory variables, one for age and one for gender.

glm(
    retained_ds ~ phone_os + age + gender,
    data = df,
    family = 'binomial'
)

This will help us understand what causes variation in retention (check the R²). The more variation we can explain, the easier it will be to identify and fix retention problems.