HEART Framework
Google’s HEART framework organizes user-experience measurement around Happiness, Engagement, Adoption, Retention, and Task success.
Download Markdown (.md)Soroush’s Product Wiki · https://wiki.soroush.uk/topic/heart-framework/
01What it means
HEART is a user-centered metrics framework created at Google for measuring experience at scale. Its five categories are Happiness, Engagement, Adoption, Retention, and Task success; teams select only the categories relevant to a product or feature.[1][2][3]
- 01HappinessAttitudes and satisfaction
- 02EngagementDepth or frequency
- 03AdoptionNew use
- 04RetentionContinued use
- 05Task successEffectiveness and efficiency
Move from a goal to signals, then a small set of metrics.
02Why it matters
Teams can collect many behavioral metrics without knowing what a good experience means. HEART pairs broad experience dimensions with a goals–signals–metrics process, helping teams connect product telemetry and research measures to intended user outcomes.
03When to use it
Use HEART when defining or revising experience metrics for a product, journey, or feature, especially when several teams need a shared measurement vocabulary. It works for both large-scale telemetry and mixed-method measurement, provided the scope and users are explicit.
04How it works
- Define the product area, user group, and decision the measurement system should support.
- Choose the HEART categories that matter; do not force a metric into every category.
- Write a user-centered goal for each selected category before choosing data.
- Identify observable signals that would indicate progress or regression toward each goal.
- Operationalize a small set of metrics from those signals, adding guardrails and qualitative evidence where numbers are incomplete.
- Review whether the metrics still represent the goal as the product and user behavior evolve.
05A practical example
For a collaborative editor, a team chooses Happiness, Adoption, and Task success. It combines perceived ease ratings, the share of eligible teams that try co-editing, completion of a shared-review task, and observed collaboration failures. It avoids using raw session duration because longer sessions may indicate either engagement or difficulty.
06Where it breaks
HEART supplies categories and a process, not universal metric definitions or target values. Engagement and retention may be inappropriate for occasional or problem-solving products. Instrumentation can miss context, and optimizing a single HEART measure can still harm accessibility, trust, or broader outcomes.
Common mistakes
- Selecting metrics before writing goals and signals.
- Measuring all five categories when only some fit the product.
- Assuming more engagement always means a better user experience.
07Sources
- Measuring the User Experience on a Large ScaleGoogle Research · Original research paperMeasuring the User Experience on a Large Scale
- Measuring the user experience on a large scaleAssociation for Computing Machinery · Peer-reviewed conference paperMeasuring the user experience on a large scale
- The HEART framework for UX metricsKerry Rodden · Creator guidanceThe HEART framework for UX metrics