Skip to content
AppDues
Log in
Contact salesSign up

Lifecycle analytics for subscription apps

If you run a mobile subscription app and want data-driven clues for improving engagement and revenue, lifecycle analytics gives you a structured way to find them. This guide covers the user lifecycle stages, the events and user properties you need to track, how to build a core funnel, and how to set up and segment retention curves so you can spot the levers that matter.

What lifecycle analytics covers

The user lifecycle is the sequence of stages a person moves through from installing your app to the end of their usage. Each stage has its own metrics, and lifecycle analytics is the practice of establishing the level of each metric and finding levers to improve it.

The stages and their focus:

  • New user: getting people to provide declarative data, finish onboarding, and grant marketing authorizations.
  • Activation: depends on your monetization model. For freemium, it centers on the core action. For pay-to-use, it includes the first purchase and the core action. With a trial period, it covers starting a trial, performing the core action, and the first purchase.
  • Engagement: core action frequency and subscription renewal.
  • Churn: reducing cancellations and expirations while increasing repurchases and reactivations.

Most of this work can be done with three chart types: funnels, retention curves, and pie charts.

Prerequisites: tooling and tracking

A handful of tools can build funnels, retention curves, and pie charts, but only a few let you segment to the granularity this work needs. The source author recommends Amplitude and Mixpanel for this analysis. Firebase may be workable, but its segmentation options are not granular enough for the author's audits. Working with your analytics team is another option, but expect many back-and-forth iterations because the work is highly exploratory.

You need two kinds of data:

  • Events: tracked actions users perform, such as completing onboarding or listening to a song.
  • User properties: personal traits such as age or nationality.

The collection of events and user properties you track is your tracking plan. A clean tracking setup is difficult to maintain, so plan for ongoing hygiene. For subscription-related events and attributes, a subscription platform can supply a clean set of data points so you can focus on the rest.

Build the core funnel

Funnels show how users move through onboarding and activation. The percentage moving from one stage to the next defines your activation performance, and the size of the drops defines your improvement potential. Start by mapping lifecycle metrics into funnels stored on a dashboard.

The core funnel gives an overview of onboarding and activation. Its goal is to show what proportion of users reach engagement and where the main friction lies. This funnel is fundamental because it can directly affect acquisition ROI, and most companies the author has worked with optimize it continuously.

Set the conversion window carefully. The conversion window is the time the analysis gives users to complete the entire funnel. Set it based on your natural usage frequency. For example, if natural usage is weekly, then performing the core action four times should happen within five weeks.

What the core funnel can reveal:

  • Onboarding performance. For free-to-use apps, a 20–40% drop between first opening the app and completing onboarding is common. At 30% or more, onboarding optimization becomes key. The bigger the drop at this stage, the stronger the impact on activation.
  • The path to the “aha” moment. The first core action should happen as early as possible. Look at the drop between the end of onboarding and the first action block. A scenario where only 60% of users who complete onboarding activate is poor; possible causes include a confusing free experience, failure to explain core value during onboarding, or friction or missing steps in the funnel.
  • Engagement target. Add core action steps until the marginal gap becomes nonsignificant. This helps you find the optimal number of repetitions for retention. CRM can then help bring users to that target.
  • Segmentation. Segment by country, persona, platform, or other relevant user properties to see how different populations pass through the funnel and uncover new friction points.

Adapt the funnel to your monetization model

Your core funnel depends on your monetization model. The free-to-use funnel differs from a pay-to-use funnel. In general, monetization models include a share of freemium and most premium experiences, so you can create one funnel per experience type:

  • Freemium funnel: aims to increase the chances users convert to a paid plan. It uses the step “Paywall Dismissed” instead of “Purchase Completed.”
  • Premium funnel: aims to make sure paid users maintain their habits with the app.

Zoom in, experiment, and repeat

Once you find the biggest drops in activation, prioritize them in a backlog. Pick one and break it into a micro-level funnel. For example, if onboarding conversion drops sharply, build a funnel with each onboarding step. Find the biggest drops, then brainstorm product and messaging experiments that address the friction. Zoom in, experiment, repeat, and improve overall engagement.

Monitor success over time

Chart conversion rate over time to see how experiments perform. This shows the conversion percentage from the funnel's last step compared with the first step over time. An upward trend is a good sign. Watch for external influences: a large influx of users from a new acquisition source can drag conversions down if traffic quality decreases.

Retention curves: overview and setup

A retention curve shows the proportion of users who return each day, week, or month after acting. It helps you understand how well you retain users over time.

Six parameters define a retention curve:

  • Start Event: triggers the analysis. Onboarding Completed is a common choice because earlier behavior is covered by funnel analysis.
  • Segment: the criterion at the moment of the start event for a user to be included.
  • Returning Event: use your app's core action. “Any active event” is common but measures opening the app rather than real engagement.
  • Frequency: set by natural usage frequency. Avoid confusing what you wish users did with their real habits. In a training app, for example, daily training may be ideal, but weekly or biweekly push-ups may be the actual habit. What counts is recurring use and subscription renewals.
  • Time Window: depends on the chart's use and how quickly the curve flattens. For an overview, use one year, or six months if the curve flattens quickly. For finding retention levers, 12 weeks is enough for daily or weekly frequency, or six months for monthly frequency.
  • Type: N-day retention counts users returning on a specific date; unbounded retention counts users returning on that date or later. N-day retention is preferred.

A note on weighted averages. Retention curves are weighted averages of each cohort in the time window and segment. The last two periods are often unreliable because only the first two cohorts are counted that long. To compensate, extend the analysis by two intervals (for example, 14 weeks instead of 12) and offset the chart by two months, so at least four cohorts reach the last interval.

What the retention overview tells you

Set up an N-day retention curve starting with Onboarding Completed, returning with the core action, at natural usage frequency, over a one-year window. The overview gives three key insights:

  • Activation conversion: already visible in the funnel, but useful for seeing how improving it affects the whole engagement scheme.
  • Rate of flattening: the core indicator of retention health. A curve that flattens means you keep a population of core users long term. A curve that keeps falling suggests ongoing user loss and possibly a lack of product/market fit. A curve that rises again at the end creates a “smile,” meaning product updates and network effects are bringing early users back.
  • Core users: the point where the curve flattens reveals your core users. Select that cohort and look at a stickiness bar chart to see how often they use the app each month. The result is your app's usage frequency.

Find actions that correlate with long-term retention

To find retention levers, segment the retention curve and compare it with the baseline of all users. The segment is users who performed a key action at least once. Use this setup:

  • Start Event: core action
  • Segment: the parameter being tracked
  • Returning Event: core action
  • Frequency: same as natural usage frequency
  • Time Window: 12 weeks, or six months if frequency is monthly
  • Type: N-day

Worked example. Suppose your movie streaming app offers features beyond watching movies: users can add films to a watchlist, watch trailers, and rate movies. Should you push those features to more users, and can you expect them to improve engagement? Segment users who used the watchlist more than once and started a movie, then compare their weekly movie-watching retention with everyone else's. If the segmented curve is higher than average, there is a correlation between the feature and high retention. The higher the curve, the more promising experiments promoting or improving the feature become.

Sizing the opportunity. Once a feature looks correlated with retention, the next question is how many users actually reach it. Funnels answer that: measure what share of your active base has ever seen a trailer, or what share of new users sees one during their first month. Put those numbers beside comparable actions, such as watchlist use. A low share means the feature is under-exposed, so messaging and onboarding changes have more room to move the needle; a high share means most of the reachable audience is already there and product improvements matter more than promotion.

Correlation is not causation. This method does not prove that the feature causes higher retention, only that the two are correlated. Maybe only highly engaged users use the watchlist because it is a niche feature, or maybe the watchlist helps users always have something to watch. Only experiments will tell.

Segment retention curves by user properties

For additional insights, repeat the same process but segment retention curves by user properties instead of events. Use the same retention curve setup, but start with Onboarding Completed. Then segment by country, platform, and person, and use pie charts to size each segment.

This method is especially useful for acquisition. If users from one country retain better than others, focusing marketing there may be worthwhile. A significant difference could also point to translation gaps or give your design team a reason to prioritize certain cultures.

Conclusion

Three chart types can significantly improve your app's engagement: funnels, retention curves, and pie charts. Funnels can achieve a lot on their own for onboarding and activation, but you need all three together to affect overall user retention. More advanced techniques exist, and each lifecycle stage requires its own expertise, but the segmentation methods here can take you to the next level.

Frequently asked questions

How is your onboarding performing?

For free-to-use apps, a 20–40% drop between first opening the app and completing onboarding is common. If you are at 30% or more, optimizing onboarding will be key to improving engagement. The bigger the drop at this stage, the stronger the impact on activation.

How is your onboarding doing in leading users to the “aha” moment?

The first core action should happen as early as possible. Look at the drop between the end of onboarding and the first action block. A scenario where only 60% of users completing onboarding activate is poor. Possible causes include a confusing free experience, failure to explain core value during onboarding, or friction or missing steps in the funnel.

What should your engagement target be?

Add core action steps in the funnel until the marginal gap becomes nonsignificant. This helps you understand the optimal number of repetitions needed for retention. CRM can then help bring users to that target.

What’s the best way to segment the funnel?

Segment by country, persona, platform, or any relevant user properties to understand how different user populations pass through the funnel. This uncovers new friction points to address.

What is the difference between N-day and unbounded retention?

N-day retention counts users returning on a specific date. Unbounded retention counts users returning on that date or later. N-day retention is generally the better choice for this analysis.

Why are the last periods of a retention curve unreliable?

Retention curves are weighted averages of each cohort in the time window and segment. The last two periods often include only the first two cohorts, so the average is less relevant. Extend the analysis by two intervals and offset the chart by two months so at least four cohorts reach the last interval.

Does a segmented retention curve prove that a feature causes better retention?

No. A higher segmented curve shows correlation, not causation. It can help you prioritize product and CRM experiments, but only experiments can show whether the feature actually drives retention.