HEART framework (Google)

There are many UX metrics - NPS, GCR, time on task, retention. How do you avoid drowning in them and pick what matters? The HEART framework is a model from Google (2010) that breaks user experience down into five dimensions: Happiness, Engagement, Adoption, Retention, Task Success.

Each dimension is a set of possible metrics. The team picks the ones relevant to its product and goals. The framework doesn't replace metrics - it organizes them. It connects to customer experience, NPS, GCR and retention. In SurveyNinja you can collect data for Happiness (satisfaction surveys) and Task Success (post-task surveys).

HEART is not about "measuring everything." It's a frame for choosing: which metrics actually matter for your case.

Definition

HEART framework - a UX metrics model developed by Google researchers (Kerry Rodden, Hilary Hutchinson, Xin Fu, 2010). It organizes the measurement of user experience into five categories: Happiness (satisfaction), Engagement, Adoption, Retention, Task Success. It helps teams choose relevant metrics instead of chaotically collecting "everything at once." It is often used together with the Goals-Signals-Metrics (GSM) process: goal → signals → metrics. It connects to feedback, surveys and analytics.

In short: five letters - five dimensions of UX. Each one has its own metrics and data sources.

H - Happiness (satisfaction)

Measures of the user's attitude toward the product: how satisfied they are, whether they recommend it, how easy it was. The sources are surveys. Metrics: NPS, CSAT, CES, SUS, store ratings. Questions: "Would you recommend the product?", "How satisfied are you?", "How easy was it to [do X]?". Happiness is a subjective slice. It complements behavioral metrics: you can use something a lot and still be dissatisfied, or the other way around.

E - Engagement

The depth and frequency of interaction. How many sessions, how much time in the product, which features people use, how often they come back. Metrics: sessions per user, time in the app, DAU/MAU, depth of feature usage. The sources are analytics and events. Engagement is not "do they like it" but "how actively do they use it." High engagement with low Happiness means the product is "sticky" but annoying. Low engagement with high Happiness may mean it's simply needed less often (a banking app used once a month).

A - Adoption

The share of users who started using the product or a new feature. New sign-ups, first-week activation, first payment, first launch of a new feature. Metrics: % activated within the first 7 days, % who tried a new feature. Adoption is "how many moved from 'aware' to 'using.'" It connects to onboarding: poor onboarding kills adoption. Conversion to activation is a typical Adoption metric.

R - Retention

The product's ability to retain users over time. Whether they return on day two, after a week, after a month. Metrics: D1/D7/D30 retention, churn rate, cohort curves. Retention is the flip side of churn. High adoption with low retention means people "come in and leave" - the product doesn't hold them. It connects to LTV: the higher the retention, the longer the customer lifetime and the higher the LTV.

T - Task Success

Efficiency and effectiveness in completing key tasks. Whether the user reached the goal, how long it took, how many errors there were. Metrics: GCR, completion rate, abandonment rate, time on task, number of errors. Task Success is "did they manage to do it." Usability tests and flow analytics are the main sources. Connection to surveys: SEQ after a task is a subjective complement to objective GCR.

Why use HEART

To structure the chaos of metrics - without a framework it's easy to gather dozens of indicators with no connection to goals. A shared language across the team - everyone understands what Happiness, Retention and Task Success mean. A focus on the user - metrics are tied to experience, not only to business KPIs. A link between design and business - HEART helps translate decisions into measurable results: "we simplified the form" → Task Success went up → conversion went up. Prioritization - you choose 2-3 dimensions for the product phase and don't spread yourself thin.

Goals-Signals-Metrics (GSM)

HEART doesn't prescribe specific metrics - it sets categories. GSM is the selection process: for each goal (Goal) you define behavioral or attitudinal signals (Signals), then measurable metrics (Metrics). Example: the goal "users easily find settings" → the signal "they find them quickly, with few errors" → the metrics "time to settings", "share who completed the task in < 30 sec". GSM helps you avoid breeding metrics "just in case" - only the ones tied to goals.

A GSM example for Happiness. Goal: users are satisfied with the checkout process. Signals: high ratings, few complaints, willingness to recommend. Metrics: CSAT after checkout, monthly NPS, share of negative open-ended responses. Without GSM it's easy to add "one more survey" with no understanding of why.

Not all five - always

Not every product or feature requires all the dimensions. A new landing page - Adoption and Task Success (did they arrive, did they complete). An existing product - Happiness, Engagement, Retention. An internal tool - Task Success and Engagement. Pick the 2-3 relevant dimensions for the context. Overloading with metrics is worse than focusing on two or three important ones.

How HEART connects to surveys

Happiness - entirely about surveys: NPS, CSAT, CES, SUS, open-ended questions. In SurveyNinja - satisfaction surveys after key actions, post-purchase, post-support.

Task Success - surveys complement analytics: "Were you able to [do X]?" - yes/no. SEQ after a task. For Adoption and Retention, surveys provide qualitative context: "Why didn't you activate?", "Why did you leave?" - exit interviews, churn surveys.

Engagement - mostly analytics, but surveys can be used for self-reported usage frequency (with caution - people are poor at assessing their own behavior).

Combining surveys and behavioral data gives the full picture: Happiness - "what they feel", Task Success - "what they do", Retention - "whether they come back".

Common mistakes

Measuring everything. Five categories are not an obligation to measure all of them. Choose 2-3 for your goals.

Metrics without goals. Is "NPS 45" good or bad? A goal and a benchmark make a metric meaningful. GSM is about linking metrics to goals.

Mixing up dimensions. Engagement and Retention are different things. The first is "how actively", the second is "whether they come back". Don't substitute one for the other.

Ignoring context. Retention for a subscription and for a one-off service are different stories. Adapt HEART to the type of product.

HEART and segmentation

HEART metrics are more useful by segment. Happiness for new users and for long-time ones is a different context. Retention by cohort shows whether retention is improving for new sign-ups. Task Success by segment - mobile vs desktop, new vs experienced. In SurveyNinja, hidden variables pass the segment (source, date, product) - when analyzing Happiness surveys you can break results down by cohort and compare.

HEART and the customer journey map

HEART metrics can be tied to the stages of the customer journey. At the awareness stage - Adoption (how many moved on to getting acquainted). At the consideration and purchase stage - Task Success (they completed, they reached the end), Happiness (how easy it was). At the usage stage - Engagement (how often they use it), Retention (whether they stay). At the support stage - Happiness (CES, CSAT). Different stages mean different metric priorities. The CJM sets the context, HEART sets which metrics to watch at each stage.

Measurement frequency

Happiness - surveys once a quarter (NPS) or after key actions (CES, CSAT). Engagement and Retention - real-time analytics, weekly/monthly reports. Adoption - on a new feature release or cohort. Task Success - on flow changes or quarterly monitoring. Not everything needs to be measured daily. Happiness and Task Success are more often collected through surveys - balance frequency against the load on users (survey fatigue).

Case: HEART for a SaaS subscription

A subscription service. They chose: Happiness (quarterly NPS, CES after a support contact), Adoption (share who subscribed in the first week after sign-up), Retention (D30, churn by cohort), Task Success (GCR for subscribing, GCR for setting up an integration). They didn't measure Engagement - the product is utilitarian, and usage frequency depends on the customer's tasks. Happiness was dropping in the "support" cohort - low CES. They went through the open-ended responses: long wait for a reply. They sped up the support response - CES went up, NPS stabilized. Task Success for setting up the integration was 60%. A usability test revealed confusion in the steps. They simplified it - GCR rose to 78%. HEART set the structure: not "measuring whatever," but "measuring by category against goals".

HEART framework - five dimensions of UX: Happiness, Engagement, Adoption, Retention, Task Success. A Google model for choosing relevant metrics. GSM links goals and metrics. In SurveyNinja - surveys for Happiness and a complement to Task Success (SEQ, NPS, CSAT).