User Engagement Score: A Health Check for Your Product

The mental model

A User Engagement Score is a single number that represents how actively and meaningfully users interact with your product. Think of it as a health indicator for your user base; a high score suggests users find value and are likely to stick around, while a low score is an early warning system for churn.

How it works

The score combines various user behaviors into one metric. A common formula is (Total Meaningful Actions / Total Users) * 100. This calculates the average number of actions per user over a period. The real power comes from weighting actions. Not all interactions are equal; creating a new document is more valuable than just logging in. For example, you might assign 3 points for 'document created' but only 1 point for 'viewed page'. This weighted approach gives a more accurate picture of true engagement than a simple count of actions.

When to use it

Use this score to track product health over time, typically on a weekly or monthly basis, to spot trends. It's also powerful for segmentation. You can compare engagement between different user groups, like users on a free plan versus a paid plan, or new users versus veterans. This helps product and customer success teams focus their efforts where they're needed most.

When not to use it

Don't use this as your only North Star metric. As a composite score, it can hide underlying issues. For instance, a high overall score might be driven by a small group of power users, masking the fact that the majority of your user base is inactive. Always use it alongside other metrics like retention cohorts, daily active users (DAU), and specific feature adoption rates for a complete view.

One canonical example

A project management platform has 1,000 total users. In one month, these users perform a total of 4,800 actions (creating tasks, adding comments, etc.). The engagement score is (4,800 Total Actions / 1,000 Total Users) * 100 = 480%. This means the average user—active or not—performed 4.8 actions that month. A more advanced version would weight 'task created' more heavily than 'comment added' to better reflect deep product usage.