Mobile Subscription
Metrics That Matter
Mobile subscription applications occupy a peculiar position in the subscription economy. They share fundamental characteristics with SaaS businesses; recurring revenue, retention-driven growth, expansion opportunities, but they operate under constraints that web-based subscriptions do not face. Where platforms (Google, Apple) control distribution, payment processing and customer relationships. Users expect lower price points than they would pay for desktop software. Competition is fierce, and switching costs are negligible. A user can delete your application and install a competitor’s in seconds.
This fundamentally changes which metrics matter, where data comes from, and how you should interpret the numbers. The frameworks designed for SaaS businesses must be adapted, to make sense in the mobile context. Understanding the distinctive characteristics of mobile subscription metrics is essential for anyone building or operating in this space.
Trial-To-Paid Conversion
In web-based subscription businesses, the friction of payment is largely under your control. You choose the payment processor, design the checkout flow, determine how many fields to require, and decide whether to ask for payment information upfront or after a trial. In mobile applications, Apple and Google control the payment experience. Users authenticate through platform credentials, confirm through platform interfaces, and receive receipts through platform systems. You have essentially no control over this flow.
This introduces friction that does not exist in web contexts. Users must have a valid payment method on file with the app store. They must navigate through platform confirmation dialogues that are deliberately designed to make users pause and consider the purchase. They must trust the platform to handle recurring billing correctly. Each of these steps introduces abandonment risk that you cannot optimise.
The consequence is that trial-to-paid conversion rates for mobile applications are typically lower than for web subscriptions. And the factors that influence conversion differ. In web subscriptions, checkout optimisation can dramatically improve conversion. Reducing form fields, adding trust signals, offering multiple payment methods: these changes can increase conversion by 10-20%. In mobile applications, you cannot optimise the payment flow without a third party plug-in like Adapty. What you can optimise is the perceived value prior to the paywall. A user must be convinced that the subscription is worth purchasing before they even encounter the platform’s payment interface.
This means that paywall positioning becomes a critical lever. Show the paywall too early, before the user has experienced enough value, and conversion suffers. Show it too late, after the user has already extracted substantial value from the free experience, and you have trained them not to pay. The optimal timing varies, but it is always a trade-off between value demonstration and conversion urgency.
Track when users encounter the paywall, how long they spend evaluating it, whether they dismiss it or proceed to the app store payment flow, and crucially, whether they return after dismissing it. Segment these behaviours by how much they have used the application, which features they have accessed, and how much time has elapsed since installation. The patterns reveal the optimal paywall strategy for different user types.
Trial-to-paid conversion should be measured not as a single aggregate rate but as a distribution over time. What percentage of users who eventually convert do so within the first day? Within the first week? Within the first month? Mobile applications often see extended evaluation periods. Users install the application, use it sporadically, and convert weeks later when they finally encounter a situation where the premium features solve an immediate need. Measuring only short-term conversion misses a substantial portion of eventual revenue.
Platform differences matter. Apple users tend to convert at higher rates than Android users, partly due to demographic differences in the user bases and partly due to differences in platform payment friction. Conversion rates also vary dramatically by geography. Users in wealthy countries with high app store penetration convert at higher rates than users in emerging markets where payment methods are less common. Aggregate conversion rates that blend these segments obscure critical dynamics that should inform pricing and go-to-market strategy.
Retention Curves
Web subscriptions tend to exhibit high early churn followed by stabilisation. Users who survive the first few months are likely to remain for extended periods. Mobile subscriptions show less predictable patterns, with churn remaining elevated even among mature cohorts.
The difference stems from the nature of mobile usage. Mobile applications are often adopted for specific, time-bound needs. A fitness application sees higher churn after New Year’s resolutions fade. A meditation application churns users who complete their intended course of practice. These patterns are not failures of product but reflections of how people use mobile applications: as tools for specific contexts rather than as permanent infrastructure. Retention should be segmented by the reason users subscribed. Users who subscribed to access a specific piece of content may churn once they have consumed it. Users who subscribed to unlock ongoing functionality may churn when they stop needing that functionality. Aggregating these different churn dynamics produces a number that is difficult to act upon.
The data for understanding retention comes from combining subscription platform data with in-application behavioural data. The subscription platform tells you when users subscribed and when they cancelled. Behavioural data tells you what they did between those events. Did they open the application regularly? Which features did they use? When did usage decline? The correlation between usage patterns and churn reveals the leading indicators that predict cancellation, enabling proactive intervention.
A particularly important metric for mobile applications is the zero-activity churn rate: the percentage of subscribers who cancel without ever opening the application during their subscription period. This is surprisingly common. Users subscribe with intent, life intervenes, they forget about the application, and months later they notice the recurring charge and cancel. This is wasted revenue from users who derived no value and might have been saved with better engagement tactics.
The flip side is highly engaged users who still churn. These are users who opened the application frequently, used core features extensively, and then suddenly stopped. This pattern suggests that they completed their goal, encountered a blocking issue, or switched to a competitor. These churns are more concerning than zero-activity churns because they represent product or positioning failures rather than mere disengagement.
Identifying and analysing these cases should be a priority for product teams.
Session Frequency & Feature Engagement
Most mobile applications are typically used in sessions: discrete periods of focused interaction separated by longer periods of inactivity. Understanding session patterns is essential for predicting retention and identifying opportunities for intervention.
Session frequency is the most basic metric: how often does a user open the application? Daily, weekly, monthly, rarely? This frequency varies by category. A meditation application might aim for daily sessions. A tax preparation application might have annual sessions. But within any category, session frequency strongly correlates with retention. Users who engage frequently are far less likely to churn than users who engage sporadically. The challenge is establishing causation. Does frequent engagement cause retention, or does retention cause frequent engagement? Likely both, which means that interventions to increase session frequency, push notifications, email reminders, streaks and incentives, can improve retention. But these interventions must be balanced against the risk of annoying users. The optimal notification strategy varies by user segment and requires experimentation.
Session duration is also revealing. Users who spend substantial time in each session are extracting value. Users who open the application briefly and close it are not finding what they need or are frustrated by the experience. Tracking average session duration and identifying drops can pinpoint where users are getting stuck or losing interest. A fitness application that sees users consistently opening the app but spending only ten seconds suggests that users are checking progress but not engaging with workouts. This signals a need to improve workout discovery or motivation.
Feature engagement goes deeper than session metrics. Not all usage is equally valuable. Some features drive retention and expansion; others are peripheral. Identifying high-value features requires correlating feature usage with subscription outcomes. Users who use feature X have half the churn rate of users who do not. Users who use feature Y are three times more likely to upgrade to a higher tier. These correlations guide product investment priorities and inform onboarding flows. This data comes from event tracking within your application. Every meaningful user action should emit an event: opened application, started workout, completed meditation, logged expense, exported report. These events accumulate in an analytics system: Firebase, Mixpanel, Amplitude and can be joined with subscription data using user identifiers. The result is a unified view of how usage patterns relate to subscription behaviour.
A particularly useful analysis is feature adoption cohorts. Segment users by whether they adopted a particular feature within their first week, first month, or never. Compare retention curves across these segments. The differences reveal which features are essential for retention and should be prioritised in onboarding. They also reveal which features are irrelevant to retention and may be candidates for removal or de-emphasis to simplify the product.
Time-to-value is another critical dimension. How long does it take a new user to experience the core value proposition? For some applications, value is immediate: open a meditation app, play a guided meditation, feel calmer. For others, value emerges over time: use a habit tracker for weeks, see progress, feel motivated. Applications with longer time-to-value face higher early churn and require more sophisticated onboarding to help users reach the point where value becomes apparent.
Expert help is only a call away. We are always happy to give advice, offer an impartial opinion and put you on the right track. Book a call with a member of our friendly team today.
Customer Lifetime Value
Calculating lifetime value for mobile applications requires the same discipline as for web subscriptions: segment by cohort, account for churn curves, incorporate expansion where it exists. But mobile introduces additional complexity around platform fees. Apple and Google take 15-30% of subscription revenue (depending on duration and programme membership), which must be deducted from gross revenue to calculate true lifetime value. A user paying £10 per month generates only £7-£8 pounds in net revenue after platform fees.
Moreover, lifetime value varies dramatically by acquisition channel. Users acquired through organic app store search tend to have higher intent and better retention than users acquired through paid advertising. Users referred by existing subscribers are often the highest-quality cohort. Understanding these differences is essential for allocation of acquisition budget. A blended lifetime value obscures the reality that some channels are profitable whilst others destroy value.
The data for calculating lifetime value comes from combining subscription platform transaction history with acquisition attribution. This is technically challenging. Mobile attribution platforms like AppsFlyer or Adjust track install sources, but connecting those installs to subscription conversions requires careful user identifier mapping. The subscription platform knows who subscribed but does not inherently know how they were acquired. Your backend must join these data sources to enable cohort-level lifetime value analysis by channel.
An often-overlooked dimension is the distinction between install-to-subscriber conversion and subscriber lifetime value. A channel might deliver high conversion rates but low lifetime value if it attracts users who subscribe impulsively and churn quickly. Another channel might deliver lower conversion but higher lifetime value if it attracts users who carefully evaluate and commit for longer periods. Optimising for conversion alone leads to acquiring the wrong customers.
Expansion revenue is less common in mobile applications than in business-to-business software-as-a-service, but it exists. Some applications offer multiple tiers or premium add-ons. Fitness applications might sell advanced training plans. Creative tools might sell additional templates or assets. Where expansion exists, it should be factored into lifetime value calculations. Users who expand are often the most engaged and the most valuable, and understanding what drives expansion informs product and pricing strategy.
In-App Purchases Versus Subscriptions
Many mobile applications offer both subscriptions and in-application purchases, either as alternative monetisation models or as complementary revenue streams. A game might offer a subscription for premium features plus in-application purchases for virtual currency. A productivity application might offer a subscription for pro features plus one-time purchases for specific templates.
The first challenge is with this accounting. Subscription revenue is recurring and relatively predictable. In-application purchase revenue is transactional and volatile. Mixing them in aggregate metrics produces misleading numbers. A spike in revenue might come from a successful in-application purchase promotion that has no bearing on underlying subscription health. A decline in revenue might be masked by strong in-application purchase performance whilst subscriptions erode. The two revenue streams must be tracked separately.
The second challenge is cannibalisation. If a user can achieve their goal through a one-time in-application purchase, why would they subscribe? If a subscriber already has access to everything, why would they make additional in-application purchases? The models must be designed to complement rather than compete. Typically, this means positioning subscriptions as providing ongoing value (features, content, support) whilst in-application purchases provide discrete value (specific items, one-time unlocks).
User behaviour differs between the models. Subscribers tend to be more engaged, as they have made an ongoing commitment. In-application purchasers are more transactional, buying when they have a specific need. Retention patterns differ. Subscribers churn through cancellation, which is a deliberate action. In-application purchasers simply stop purchasing, which is passive disengagement. The interventions to retain each type of user are different.
From a metrics perspective, you need parallel frameworks. For subscriptions, track monthly recurring revenue, churn, net revenue retention, and lifetime value. For in-application purchases, track average revenue per user, purchase frequency, repeat purchase rates, and customer lifetime value calculated from transaction history rather than recurring amounts. Mixing these metrics produces confusion.
The data infrastructure must accommodate both models. Subscription platforms like RevenueCat handle subscriptions but also track in-application purchases, providing a unified view. Your analytics must segment users into cohorts: subscribers only, in-application purchasers only, users who do both. The behaviours and economics of each cohort are distinct and should inform different strategies.
A particularly valuable analysis is conversion between models. What percentage of in-application purchasers eventually subscribe? What percentage of subscribers make additional in-application purchases? Do in-application purchases serve as a gateway to subscription adoption, or do they satisfy the need and prevent subscription conversion? The answers vary by application and pricing structure, but understanding these dynamics is essential for optimising the monetisation mix.
Platform-Specific Metrics
Mobile subscription applications typically operate across both iOS and Android, and the platforms exhibit different characteristics that must be measured and optimised separately. iOS users generally have higher willingness to pay, leading to higher conversion rates and higher price tolerance. Android users are more price-sensitive but represent a larger global market, particularly in emerging economies.
These differences are not merely averages; they reflect fundamentally different user bases with different expectations and behaviours. An application might charge £10 per month on iOS and £7 on Android, reflecting the willingness-to-pay differential. Conversion rates might be 40% on iOS and 35% percent on Android. Lifetime value might be twice as high on iOS. These disparities mean that metrics must be segmented by platform, not blended.
The challenge is that data often lives in platform-specific silos. Apple provides analytics through App Store Connect. Google provides analytics through Google Play Console. Subscription platforms provide unified views, but even they may show platform differences in transaction formats and metadata. Stitching together a coherent cross-platform view requires deliberate data architecture.
Platform policies differ in ways that affect metrics. Apple’s privacy changes have made attribution more difficult on iOS, reducing the visibility into which acquisition channels drive installs and conversions. Google has historically provided richer attribution data. This asymmetry means that your confidence in acquisition analytics may differ by platform, and strategies that work on one platform may not be measurable on the other.
The practical implication is that product and pricing decisions may need to be platform-specific. A feature that drives retention on iOS might not have the same effect on Android. A price point that works in the United States might be prohibitive in Brazil. The temptation is to maintain parity across platforms for simplicity, but doing so leaves value on the table. Platform-specific optimisation is more complex but more effective.
One Source Of Truth
The mobile subscription metrics that matter do not start in a dashboard, they start in a data dictionary. Before you track anything, define it: what counts as an “active subscriber” on iOS vs Google Play, how you treat grace periods, refunds, upgrades/downgrades, trials, family sharing, and cross-device restores. Write down the exact logic, the event sources, and the owner for each metric. This is how you prevent the most common failure mode in subscription reporting: different teams making different decisions from different definitions of the “same” number.
Next, centralise the raw facts in a data warehouse. Pipe in store-side signals (receipts, purchase tokens, server notifications), your backend entitlements, and product analytics events, then join them around stable keys (user, device, subscription, transaction) with clear lineage. The warehouse becomes your ground truth, not the app store console, not a third-party dashboard, and not a spreadsheet. Once the inputs live in one place, you can audit gaps, reconcile discrepancies, and trust that every stakeholder is looking at the same underlying story.
Finally, model the metrics so they can be automated and decision-aligned. Build a canonical subscription model (current state, plan, renewal date, lifecycle stage), and layer decision-facing outputs on top: conversion by acquisition channel, trial-to-paid by paywall variant, churn by tenure band, LTV by cohort, and upgrade propensity by feature usage. Automate them with scheduled transformations and tests, and publish them as “gold” tables or semantic metrics that power dashboards, alerts, and lifecycle triggers. The goal is not to measure everything, it’s to reliably measure what changes your next action: what to ship, what to price, who to target, and where to invest.
If you need help in this area, why not talk to the friendly team at 173tech who have done this many times before?
Get In Touch
Our friendly team are always on hand to answer questions, troubleshoot problems and point you in the right direction.
