HEART: metrics for user-centered products at scale
Five metric families — Happiness, Engagement, Adoption, Retention, Task success — operationalized through a goals → signals → metrics process.
Research question
How do you measure user experience at scale, beyond business KPIs that say nothing about experience quality?
The framework
Happiness (attitude: surveys, ratings), Engagement (depth/frequency of use), Adoption (new users of a product or feature), Retention (returning users), Task success (completion, time, errors). Not every product needs all five — pick what matches the product’s goals.
The lasting contribution is the goals → signals → metrics discipline: articulate the goal in words, choose observable signals of it, only then define metrics. It blocks the classic failure of measuring whatever the analytics tool shows by default.
What it means in practice
For any redesign, write one HEART row before shipping: goal (“users find settings faster”), signal (time to reach setting, support tickets), metric (median seconds, tickets/week). That row is your success definition — agreed before the data arrives.