Joy Foroughi
Executive Assistant
App Teams: Find the Biggest Paid Conversion Leak in 90 Minutes
When an app gets downloads but few paying subscribers, the problem may sit in onboarding, paywall exposure, the trial experience, or billing. Start by checking your first-session tracking and comparing conversion by acquisition source and platform. This 90-minute review helps you identify a likely bottleneck and choose one experiment, provided you already have reliable tracking and enough historical data.
When an app gets downloads but few paying subscribers, the problem may sit in onboarding, paywall exposure, the trial experience, or billing. Start by checking your first-session tracking and comparing conversion by acquisition source and platform. This 90-minute review helps you identify a likely bottleneck and choose one experiment, provided you already have reliable tracking and enough historical data.
TL;DR:
RevenueCat’s 2025 report notes that 82% of trial starts occur on the day of download, so measure first session activation and paywall exposure before changing prices.
Low download-to-trial rates can signal onboarding, offer visibility, eligibility, or traffic-quality issues; low trial-to-paid rates warrant investigation of value, trial design, and billing.
RevenueCat’s 2026 report gives a 2.6% North American median D35 conversion and medians of 10.7% for hard paywalls versus 2.1% for freemium. These observational benchmarks are not universal targets.
Trials lasting 17–32 days had a 42.5% median conversion rate, versus 25.5% for four days or less; test duration against lifetime value.
For a new Android purchase, grant access after the transaction completes and verification confirms entitlement; include refunds, chargebacks, and churn in experiment results.
How do you find your biggest funnel leak in 90 minutes?
Use this review before committing to pricing or design changes. The goal is one evidence-backed hypothesis and one experiment to investigate it.
This review assumes reliable tracking and enough historical data. If events are missing, use the session to identify measurement gaps. You cannot add tracking today and immediately measure a 35-day conversion outcome.
Time | Action | Output |
|---|---|---|
0–15 minutes | Check first opens, activation events, paywall views, trial starts, checkout starts, verified purchases, and payment errors. | A list of usable events and tracking gaps. |
15–35 minutes | Compare funnel steps using cohorts with complete observation windows. Separate trial subscriptions from direct-to-paid purchases. | Baseline conversion rates with clear denominators. |
35–55 minutes | Split results by platform and acquisition source. Check first-session paywall reach and avoid drawing conclusions from tiny segments. | One or two segments worth investigating. |
55–75 minutes | Inspect likely causes, including onboarding friction, offer eligibility, pending transactions, and payment failures. | A ranked shortlist of plausible opportunities. |
75–90 minutes | Choose one hypothesis, experiment, primary metric, guardrails, and owner. | A concrete next step for the team. |
Prioritize steps by the additional paying users a realistic improvement could produce, supported by evidence and weighed against implementation effort. The largest raw drop-off is not automatically the best opportunity.
Pro Tip: Estimate the users entering a step × a plausible improvement in its conversion rate × the downstream chance of becoming a paying subscriber. Use this as a rough prioritization aid, not a forecast or proof of causation.
RevenueCat's 2025 report finds that 80–90% of trials begin on Day 0, which makes first-session tracking a useful place to start. A practical diagnosis splits the funnel into install→activation, activation→paywall view, paywall view→trial or purchase, trial→paid, and renewal. These stages help narrow the investigation to possible causes such as promise mismatch, onboarding friction, offer configuration, perceived value, or billing problems. Include a separate path for users who subscribe without a trial.
What do D35, D30, and trial-to-paid actually tell you?
These are cohort metrics, and the denominator decides what they mean. Mixing them hides the real leak.
D35 download→paid is the share of a download cohort that makes its first paid subscription payment within 35 days. Compare only cohorts with a full 35-day observation window. RevenueCat’s 2026 report gives a North American median of 2.6%.
D30 download→trial is the share of a download cohort that starts a trial within 30 days. It can help assess trial uptake, but it does not isolate the first session. Measure Day-0 trial starts and first-session paywall reach separately.
Trial→paid is the share of trial starters who become paying subscribers within a defined follow-up window. Include only trials that have had time to end and payments to resolve; do not compare completed trials with trials still in progress.
Segments help explain differences. Start with platform and acquisition source, then examine category, country, price point, or behavior where sample sizes allow meaningful comparisons.
Use a consistent denominator. Store downloads, first opens, and new users recorded by your analytics system are different measures. If your data starts at first open, label the metric first-open-to-paid rather than comparing it directly with download benchmarks. Count unique users consistently and report direct-to-paid purchases separately from trial conversions.
Strong trial uptake with weak trial-to-paid conversion warrants investigation of the trial experience, offer expectations, and billing. Weak trial uptake may reflect onboarding, paywall reach, eligibility, traffic quality, or tracking gaps. These patterns suggest hypotheses; they do not establish causes. Compare benchmarks only when the category, business model, denominator, and observation window are sufficiently similar.
Which onboarding changes lift Day-0 activation and trial starts?
Users need a clear reason to subscribe. Depending on the product, that value may be demonstrated through a useful action, an onboarding preview, or clear examples before the paywall.
Map the shortest path to your "aha" moment and cut every step that doesn't serve it. Apple's subscription guidance notes that long or complex sign-up processes lower conversion.
Where sign-in is necessary, consider Apple or Google sign-in to reduce friction. Ask only for information needed at that stage.
Test paywall timing against your product’s path to value. Some apps can demonstrate value before asking users to subscribe; others need to communicate it through onboarding or a preview.
Help users experience a relevant premium benefit with minimal setup. Measure whether that experience improves activation and paid conversion.
Consider a reverse trial, which gives temporary premium access without upfront payment details, when users can experience value that way. Measure paid conversion and revenue as well as engagement.
Pro Tip: Log the timestamp of a clearly defined activation event. Compare it with paywall timing to identify a hypothesis worth testing; an early paywall alone does not establish a problem.
A 2026 study in the Journal of the Academy of Marketing Science found that habits formed in a free version can either help or suppress conversion. Treat this as a reason to investigate how free and premium usage relate to conversion in your own product, rather than assuming that more free usage always helps. For first-run flow patterns, see our guide to mobile app onboarding that survives Day 7.
How should you design paywalls and trial offers?
Paywall design is a set of trade-offs, so test rather than copy. In RevenueCat’s 2026 report, median D35 download-to-paid conversion was 10.7% for hard-paywall apps versus 2.1% for freemium apps. That is an observational comparison, not a universal case for hard paywalls. Hard paywalls concentrate conversions early, while freemium spreads them over time.
Hard paywall vs. freemium: a hard paywall may suit products whose value can be communicated upfront; freemium may suit products where value builds with use. Test the choice against paid conversion, retention, and revenue rather than treating either model as a default.
Trial length: RevenueCat’s 2026 report gives a 42.5% median trial-to-paid rate for 17–32 day trials versus 25.5% for trials of four days or less. That is correlation, so A/B test durations against LTV.
Placement: test relevant offers at moments when users understand the benefit. Track dismissals and abandonment alongside conversion rather than simply increasing exposure count.
Offer mechanics: show the trial duration, amount due, renewal price, and billing frequency clearly. Check eligibility, supported introductory offers, and plan-switch rules for each store.
We cover layouts and copy in our paywall design patterns guide.
How do pricing and packaging affect conversion?
Price works as a filter, not a universal fix. RevenueCat's analysis shows higher-priced apps often convert better from download to trial, but execution matters more than price alone.
Use clear plan names and explain which benefits each plan includes. Use App Store Connect’s documentation to configure subscription groups, levels, and plan changes correctly.
Test eligible introductory pricing against a free trial where the stores support it. Compare paid conversion, net revenue, refunds, and lifetime value (LTV), rather than trial starts alone.
Use time-to-value to form a trial-length hypothesis, then test it. A longer trial may give users more time to experience value, but can also delay revenue and increase service costs.
Consider a different plan structure only when it addresses a specific customer need or a measured conversion problem.
Are your conversions real? Checking billing and entitlements
A client-side purchase event alone does not establish a valid entitlement or a successful paid charge. For new Android purchases, Google’s billing integration documentation separates purchase verification, granting access, and acknowledging delivery. Do not grant access for an initial purchase while it remains pending. Verify the completed purchase through your backend or a trusted billing service and acknowledge it as required. Existing access should follow the subscription’s current entitlement state.
Handle pending transactions and linked purchase tokens correctly, following the Play Billing subscription docs on replacement and proration modes.
Use store recovery tools: Google's Dynamic Grace Period and Retention Offers, and App Store billing grace periods.
Track trial access, successful paid charges, and active entitlements separately. A valid free-trial entitlement is not a paid conversion. Include refunds, chargebacks, service costs, and churn when assessing the business result.
Instrument payment error rates, checkout abandonment, and time from checkout start to completion.
RevenueCat's Android analysis found similar trial-to-paid medians: 32.5% on Android and 32.6% on iOS. The analysis also notes that Android trial starters may be a more selected, higher-intent group. These aggregate results make trial entry worth investigating, but do not establish the cause of a gap in your own app. Validate your active Play offer configuration before changing price. Check renewal and payment-recovery outcomes after fixing the initial conversion path.
How do you show the right offer to the right user?
Segmentation lets you test where it counts without bloating your experiment queue.
Use source, campaign, platform, and pre-experiment behavior to identify a clear test audience. Keep a randomized control within that audience and avoid generalizing the result to all users.
Use a short onboarding quiz or behavioral signals to pick a tailored paywall and offer.
Start with simple contextual timing based on relevant screens or actions. Consider more complex targeting only when your traffic and evidence justify it, and validate against retention and revenue.
Consider hybrid monetization, pairing subscriptions with consumables, to capture value from users who never subscribe.
What testing roadmap keeps you from false wins?
Test one clearly defined hypothesis at a time within the affected funnel and audience.
Check tracking and establish a baseline before launching the test.
Use a small number of user interviews or appropriately configured session recordings to explore likely causes. Treat these observations as qualitative evidence, not a conversion estimate.
Randomize eligible users and choose one primary outcome tied to the hypothesis, such as D35 paid conversion per eligible new user. Keep the eligibility rule and denominator consistent across variants. Use trial starts and trial-to-paid as diagnostic metrics, with retention, refunds, and net revenue as guardrails.
Define the sample requirement, minimum effect worth detecting, and observation window before launch. Let each cohort complete the relevant follow-up period; changes to trial duration require particular care because payment occurs at different times.
Evaluate the primary outcome alongside guardrails. Use a defined later window to monitor renewals, refunds, chargebacks, and realized revenue before making broader LTV claims.
Set stopping rules before launch. Serious billing failures, crashes, access problems, or material deterioration in guardrails may justify a rollback. Do not declare a winner simply because an early result looks positive. Low-traffic apps may need more time than the proposed diagnostic session to reach a reliable experimental conclusion.
What we've seen: diagnosis beats redesign
A paywall redesign may miss the underlying problem if users never reach the offer or encounter billing failures. Start by identifying where users drop out and testing the likely cause before investing in a larger redesign.
Keep responsibilities clear: product owns the success metric, engineering owns event quality and purchase verification, and design owns the onboarding and paywall experience. Bring in outside help when billing fixes are complex, the design scope is large, or the team lacks capacity to implement and measure the changes.
How we can help you convert more downloads
TouchZen helps app teams connect product strategy, UX/UI design, and implementation. Start with a focused review of your onboarding, paywall, and billing journey, then use the findings to prioritize changes and define how success will be measured. Contact TouchZen to discuss your conversion funnel.
FAQ
Why does my app get downloads but few paying subscribers?
Users may drop out before activation, miss the subscription offer, find the trial unconvincing, or encounter payment problems. Acquisition quality and tracking gaps can also affect the numbers. Compare each funnel step using consistent cohorts, then investigate the most plausible opportunity before changing pricing or design.
How can billing errors affect subscription conversion?
Pending purchases, failed payments, offer eligibility issues, or delayed entitlement updates can interrupt the path to a subscription. Track checkout errors and verified paid charges separately from trial access. Confirm that purchase verification and access handling work correctly before interpreting a weak conversion rate as a product-value problem.
What is a good download-to-paid conversion rate?
It depends on category, platform, and model, so compare like with like. RevenueCat's 2026 report frames this as a cohort metric, and the 2.6% North America median it reports is a starting reference, not a target.
Should I switch from freemium to a hard paywall?
Not on benchmarks alone. Hard paywalls convert earlier in RevenueCat's data, but freemium spreads conversions over time and suits products where value builds with use. Test both against retention and LTV before committing.
Sources
Apple — Offer auto-renewable subscriptions in App Store Connect
Google — Driving growth on Google Play: The next era of subscriptions
