Start with the single fact that reframes everything: on mobile, a free trial is opt-out. Apple and Google auto-convert the trial into a paid subscription unless the user actively cancels. Web and SaaS trials are usually opt-in — the user has to come back and choose to pay. That default flip is enormous. Across web and SaaS, opt-out trials convert around 48.8% versus roughly 18.2% for opt-in — the same product, a different default, nearly triple the conversion. It is why mobile trial economics look nothing like the web advice you may have read.
Treat those numbers as benchmarks, not laws. They come from RevenueCat and Adapty’s 2026 reports plus web/SaaS sources, which sample different apps with different definitions — so use them to find your gap, then test your way forward. For the full revenue picture beyond trials, see the pillar playbook.
1. Mobile trials are opt-out — design around it
Because the store renews automatically, the mobile trial’s job is not to *re-sell* the user at the end — it is to keep them from cancelling before the first charge. Global mobile trial-to-paid sits around 25–37% depending on whose dataset you read: RevenueCat reports a 5–9 day trial converting at 37.4%, while Adapty’s global average across all lengths is 25.6%. Both are far above the ~18.2% web opt-in figure. The practical implication: your trial is a retention problem disguised as a conversion problem.
- Make the auto-renew terms honest and visible on the paywall. Surprise charges drive refunds and one-star reviews; clarity actually lifts trust-driven conversion.
- Instrument the first session like it is the whole game — because it nearly is (see day zero below).
- Send a pre-expiry reminder (benefit + time remaining). It is allowed, it is expected, and it converts.
- Remember the opt-out advantage cuts both ways: a confused or unengaged trialist who forgets to cancel becomes an immediate refund and churn risk, not a happy payer.
2. Win day zero
The most decisive moment in the entire trial is the first day. 80–90% of all trial starts happen on day zero — the same session as the install. If a user does not start the trial on day one, they almost never come back to start it later. That collapses the problem: most of your trial volume is won or lost in the first few minutes, right after the user feels the app’s value for the first time.
- Trigger the trial offer right after the aha moment — the first time the app obviously delivers value — not on a cold first launch.
- Then test moving it earlier, even before the aha moment, kept dismissible: this is the same early-paywall pattern that lifts sign-ups dramatically. See paywall optimization.
- Reduce friction to start: a free-trial toggle and a single tap beat a multi-step sign-up.
- Personalize the offer from onboarding answers so the first thing the user sees maps to why they downloaded.
3. The trial-length debate — where the data disagrees
Here the two biggest providers genuinely disagree, and you should know it before you copy anyone. Adapty finds a 5–9 day sweet spot for most apps (about 52% of apps use it). RevenueCat finds the opposite tail wins: longer trials of 17–32 days convert ~70% better than the shortest ones. Both are right for their samples — which is exactly why there is no universal answer. The honest move is to test trial length on your own users rather than inherit someone else’s optimum.
| Trial length | Trial→paid (RevenueCat) | Pattern |
|---|---|---|
| ≤4 days | 25.5% | Shortest trials convert worst on this dataset |
| 5–9 days | 37.4% | Adapty’s reported sweet spot (~52% of apps use it) |
| 10–16 days | 35.4% | Mid-range, slightly below the 5–9 band |
| 17–32 days | 42.5% | RevenueCat: long trials ~70% better than shortest |
Conversion by category matters as much as length: Travel converts around 43.5%, Health & Fitness 37.7%, Education 33.2%, Media & Entertainment 27.6%, and Gaming 25.0% (RevenueCat 2026 medians). A short trial can be right for an impulse category and wrong for a considered one — context, not dogma.
4. Trial reminders and cancellation timing
A large slice of trial loss is simply people cancelling on the very first day — and the rate is tightly linked to trial length. On a 3-day trial, 55.4% of cancellations happen on day zero; on a 30-day trial only 31.1% cancel on day zero. A short trial pressures the user to decide before they have built a habit, so more of them bail immediately; a longer trial gives habit time to form before the question is even asked.
| Trial length | Cancel on day 0 | Reading |
|---|---|---|
| 3 days | 55.4% | Decision forced before habit forms — most bail immediately |
| 7 days | 39.8% | Some breathing room, still front-loaded cancels |
| 14 days | 35.7% | Habit has time to start before the question lands |
| 30 days | 31.1% | Lowest day-zero cancel; habit usually formed by expiry |
- Send a reminder before expiry — with the concrete benefit they will lose and the time remaining. It is expected on mobile and consistently converts.
- Use the first session to deliver an obvious win, so the day-zero cancel instinct never fires.
- For longer trials, add a mid-trial nudge that resurfaces value, not just a final warning.
5. Feature gating is 90% of the outcome
For freemium especially, what you gate is roughly 90% of success. The free tier must let the user complete the basic job well enough to feel value; premium must unlock the high-value layer they will keep coming back for. Gate too little and no one upgrades; gate the core job itself and no one stays long enough to want premium. The line between "useful free" and "compelling premium" is the entire freemium business.
- Free = the basic job done well. Premium = the high-value layer (volume, quality, speed, automation, or removal of a real limit).
- Gate on a dimension the engaged user hits naturally — usage limits, export, advanced output — not on the first thing they touch.
- A free-trial toggle on the paywall reframes premium as "try the high-value layer risk-free," which is a recurring redesign winner.
- For pricing the layers you gate, see subscription pricing strategy.
6. Trial vs hard paywall vs freemium
These are not interchangeable. A hard paywall (pay before any real access) posts a median download-to-paid around 10.7%, versus 2.1% for mobile freemium — but that ~5× gap mostly reflects filtering of low-intent users, not better persuasion. General freemium sits in a 1–10% range depending on definition; on mobile the realistic figure is closer to that 2.1%. The decisive long-term advantage of trials is downstream: trial-then-paid users retain 1.4–1.7× better than people who buy directly, so a trial is a retention investment, not just a conversion tactic.
Hard paywall
~10.7% download-to-paid, but it filters out low-intent users. Best for high-intent traffic (search, ASO). Highest headline number, smallest top of funnel.
Free trial
Opt-out on mobile (~25–37% trial-to-paid). Lets users feel value first; trial-then-paid retains 1.4–1.7× better. The default mobile choice for considered purchases.
Freemium
~2.1% mobile download-to-paid (1–10% range generally). Largest top of funnel, lowest conversion; success lives or dies on feature gating.
7. Retention is the real scoreboard
Conversion gets the headline; retention pays the bills. Among trial subscribers, day-380 retention splits hard by plan length: annual 19.9%, monthly 14.2%, weekly 5.5%. A weekly trial can post a flashy initial conversion and still leave you with almost no one a year later, while an annual trial converts fewer people but keeps far more of them. Always read trial performance through the retention lens, not just the day-one conversion number.
- Pick the metric that matters: trial-to-paid is the start, but day-180 and day-380 retention decide the lifetime value.
- Pair longer plan durations with the trial when retention is the goal — annual trials retain far better at day 380.
- Run trial experiments long enough to see at least one renewal cycle (often 4–8 weeks) before declaring a winner.
Where AI pricing fits
Once your trial flow, day-zero, length, and gating are solid, the next frontier is personalization — meeting each user closer to their actual willingness to pay instead of one trial-and-price for everyone. Because the stores forbid showing arbitrary prices, this is done through eligibility-aware offers rather than raw price changes. This is the layer Monetai operates in: it predicts each user’s purchase intent and serves a personalized trial or discount only to users who need one, so you capture incremental conversions without giving away revenue to people who would have paid anyway. It sits on top of whatever paywall and trial you already run.
Want to see where your own trial pricing lands against the market? Browse live app pricing benchmarks from App Pricing Lab’s daily crawl of 135,000+ apps, or start from the playbook hub.