Benchmark · Subscription apps
What the funnel numbers should look like.
Everything below is attributed to a named published source with a date and a sample size. Where a widely quoted figure is misleading, we say so instead of repeating it.
- Last checked
- September 2026
- Sources
- 4 published studies
- Our own data
- Not included
The finding that matters more than any median
RevenueCat's State of Subscription Apps 2026, built on more than 115,000 apps and over $16 billion in tracked revenue, reports that the top quartile of subscription apps grew monthly recurring revenue by more than 80% year on year, while the bottom quartile shrank by more than 33%.
That dispersion is the real headline. It means a category median is describing a distribution so wide that the middle of it is not a meaningful target for anyone. Two apps in the same category, at the same price, with the same acquisition cost, routinely sit at opposite ends of that range – and the difference is almost never media buying skill.
Pricing behaves differently than people expect
Airbridge's 2026 subscription pricing analysis, built on roughly $3 billion in subscription revenue, found median monthly prices differ sharply by category – health and fitness around $9.70 against gaming near $4.99 – and, more usefully, that higher-priced apps converted downloads to paid at a better rate: around 2.7% at day 35 for high-priced apps against 1.5% for low-priced ones.
The intuitive model, where cheaper converts better, is wrong at the population level. Price signals seriousness, filters for intent, and correlates with products that solve a problem worth paying for.
| Figure | Value | Source | Population |
|---|---|---|---|
| Top-quartile MRR growth | +80% YoY | RevenueCat, State of Subscription Apps 2026 | 115,000+ apps, $16bn+ revenue |
| Bottom-quartile MRR growth | −33% YoY | RevenueCat, 2026 | Same |
| Median monthly price, health & fitness | $9.70 | Airbridge, 2026 pricing benchmark | ~$3bn subscription revenue |
| Median monthly price, gaming | $4.99 | Airbridge, 2026 | Same |
| D35 download-to-paid, high-priced apps | 2.7% | Airbridge, 2026 | Same |
| D35 download-to-paid, low-priced apps | 1.5% | Airbridge, 2026 | Same |
Source: Figures as published by each provider. Verify against the current edition before quoting – these reports are revised annually and the samples change.
What none of these sources can tell you
Your trial-to-paid rate. It is governed by trial length and payment gating, which vary so much between apps that a cross-company median is close to meaningless. A three-day trial with card details up front and a seven-day trial without them are different products, measured by the same word.
Your payback period. Nobody publishes it, because it depends on acquisition cost, which is specific to your channels, your market and your creative. It is also the only number that answers the question you actually have.
Whether your churn is bad. Churn measured in month one is always the worst it will ever be, and reports differ on which month they mean. Comparing your month-one figure to someone's blended annual number will convince you that you have a crisis you may not have.
How to build a comparison that means something
Compare cohorts to your own earlier cohorts, segmented by acquisition source. Same product, same price, same audience – so a difference in the number is a difference in something you changed, which is the only kind of comparison that tells you what to do next.
Use industry figures for one thing only: checking whether you are in an entirely different league. If your trial-to-paid is a third of everything published, something is broken. If it is close, the median has nothing more to offer you.
Questions people actually ask
- What is a good trial-to-paid conversion rate?
- The question is not answerable across companies, because the number depends almost entirely on trial length and whether payment details are required up front. A three-day trial with card details captured converts at a multiple of a seven-day trial without them – and produces far fewer trials. Compare yourself to your own previous cohorts instead.
- Why do published benchmarks vary so much?
- Because each one measures a different population. A subscription platform's data covers apps that use that platform. An attribution provider's data covers apps that pay for attribution – a larger, better-funded set than average. Neither is wrong; both describe a specific slice.
- Which single number matters most?
- Payback period, and it is not usually published because it depends on acquisition cost, which is company-specific. It is the only figure that tells you whether you can spend more tomorrow, which is the actual decision.
- How much does price affect conversion?
- Less than most teams expect on the trial-start step and more than they expect on retention. Airbridge's 2026 analysis found higher-priced apps converted downloads to paid at a better rate than lower-priced ones – the opposite of the intuition that cheaper converts better.
- Should we compare against category medians at all?
- Only to check whether you are in a different league entirely. If your trial-to-paid is a third of what everyone reports, something is broken and worth investigating. If it is within striking distance, the median tells you nothing about what to change next.
Turn a benchmark into a decision.
The payback calculator takes your own numbers and tells you the month a cohort becomes profitable. That is the figure worth acting on.