Maybe LabsMaybe Labs
← All posts
Guide·5 min read

Feature Adoption Metrics: Measure What Happens After an Update

Maybe LabsBy the Maybe Labs team

Feature adoption metrics show whether eligible users try a capability, complete a useful task, and return to it. An announcement view is an exposure signal; successful feature use is a different event. Start by defining who could use the feature and what completed use means before choosing a dashboard.

This guide uses a fictional saved-view release. All counts are illustrative, not Maybe Labs results or industry benchmarks. The feature lets a user save status and owner filters as a personal view, then reopen that view. The useful outcome is returning to a saved workflow, rather than merely opening the save dialog.

Define the eligible population before calculating adoption

Write a measurement contract: the feature, unit of analysis, eligibility rule, rollout date, observation window, and success event. For this example, the unit is a signed-in human user in an enabled workspace. Exclude staff and automated test accounts. Include users who can access the feature even if they never saw the announcement; otherwise the headline adoption rate silently becomes a campaign conversion rate.

For an account-level feature, use eligible workspaces instead. Do not divide unique workspaces by users. Freeze the cohort when comparing launch weeks, or report changing eligibility explicitly. A staged rollout can expand the denominator while usage grows, making a falling percentage look like a product regression.

MetricDefinition for this exampleQuestion answered
Announcement reachUnique eligible users with a recorded announcement exposure / eligible usersHow much of the available audience saw the communication?
First-use adoptionUnique eligible users who successfully saved a first view / eligible usersWho completed an initial useful action?
View-to-use conversionExposed eligible users who then saved a view within seven days / exposed eligible usersWhat happened after a measured exposure?
Repeat-use rateFirst-time savers who reopened a saved view on a later day within 14 days / first-time savers with a full 14-day windowDid initial use become a returning behavior?
Time to first useElapsed time from eligibility to the first successful saveHow long does discovery and first success take?

A worked example with honest denominators

Suppose 200 users became eligible on October 1. Within the first seven days, 120 saw a measurable announcement exposure, 50 saved their first view, and 36 of those 50 saved after an exposure. After a complete 14-day follow-up, 30 of the 50 savers reopened a view on a later day. The 14 other savers found the feature through some other route, or their exposure was not recorded.

CalculationResultInterpretation
120 / 20060% reachThe exposure event reached 60% of the eligible cohort.
50 / 20025% first-use adoptionOne quarter of the cohort completed a first save.
36 / 12030% view-to-use conversionThirty percent of measured viewers subsequently saved.
30 / 5060% repeat useThree fifths of first-time savers returned on a later day.

These percentages answer different questions. The 30% campaign funnel cannot replace the 25% feature-adoption rate. Nor does it show that the announcement caused those saves: users who watch a feature video may already be more interested or active. Label the dashboard with both numerator and denominator so a reader can reconstruct each percentage.

Use an event plan that records successful actions

EventWhen to send itUseful properties
feature_eligibleA user enters the defined rollout cohortfeature_key, rollout_version, eligible_since
update_announcement_seenThe chosen surface actually records a qualifying exposurecampaign_id, surface, feature_key
personal_view_savedThe application confirms that a view was persistedview_id, feature_key, rollout_version
personal_view_openedThe saved configuration loads successfullyview_id, feature_key

Use stable event names and a shared feature_key. Treat identity consistently across marketing and product surfaces. A video player starting, an email thumbnail click, and an in-app message exposure are different observations; record them separately if they answer different questions. Keep private record contents and personal data out of event properties unless they are needed and handled appropriately.

Build the exposure-to-use funnel in PostHog

In PostHog Funnels, create steps for your exposure event and successful save event. Apply the eligibility filter, choose a seven-day conversion window for this worked example, and inspect the people who drop between steps. The window is a planning choice, not a universal recommendation. Save the report with its date range and filters.

Build the all-eligible first-use metric separately, because an exposure-first funnel omits people who discover the feature elsewhere. For repeat use, examine a later-day reopen among first-time savers and exclude cohorts that have not yet had their complete follow-up period. Compare like-for-like cohorts rather than treating a partially observed launch week as a complete one.

Use Appcues to distinguish reach from adoption

Appcues dashboard guidance distinguishes feature-launch reach, first use, and ongoing usage. Map your announcement and successful-use events to those questions. Its feature-announcement workflow guide describes following up according to whether someone has used the feature. Configure that logic around your actual success event, not a generic click.

Decide what to improve from the pattern

  • Low reach, healthy use among viewers: inspect placement and audience targeting before rebuilding the feature explanation.
  • Healthy reach, weak first use: check whether the announcement explains the entry point, eligibility, and practical benefit. Then inspect friction in the task itself.
  • Healthy first use, weak repeat use: investigate whether the saved workflow is useful on a later day. More announcements may not solve a weak recurring need.
  • Many clicks but few successful saves: inspect errors, permissions, and naming friction in the product.

For a product update video, pair the clip with the entry point and complete written release note. Use a descriptive campaign link where appropriate, but keep a view metric separate from a product event. Our video marketing metrics guide covers attention measures; product video ROI explains the broader evaluation.

Do not turn an observational funnel into an uplift claim

If you need a causal estimate, plan a suitable experiment with a valid comparison group and a clear success metric. Account for rollout eligibility, exposure contamination, sample size, and the time users need to return. A higher adoption rate among viewers is useful descriptive evidence, but it is not by itself proof of incremental lift. Report the absolute counts alongside uncertainty instead of presenting a small cohort as a universal benchmark.

What is a good feature adoption rate?

There is no universal threshold in this guide. The relevant baseline depends on eligibility, task frequency, product maturity, and the observation window. Compare a clearly defined metric with your own prior cohorts.

Should adoption count a button click?

Count the completed outcome when possible. Opening Save view shows intent; a confirmed persisted view is stronger evidence of initial use.

How is repeat use different from retention?

Repeat feature use describes a specific returning action. Product retention can be defined around a broader activity and should be measured separately.

Make your next launch in motion

Bring a brief, brand assets, and product captures. Review the story, visuals, and exported video before publishing. Request free private-beta access.

Get early access →

Keep reading