How long should a free trial be, and which paywall converts best?
By gr0w Research · Updated · From 185 founder interviews · 6 min read
Short answer
Among founders in an analysis of 185 interviews (42 of which name paywall or pricing as a growth driver), the two who A/B tested trial length both landed on 7 days over 3 days. Most who tested a hard paywall right after onboarding reported higher conversion than freemium, such as a 50% lift for Glam Up/Sprout's founder and about 25% more revenue for Stronger. Annual plans with the trial offered as an option, plus a trial-reminder screen, were the most-cited ways to raise revenue. Results are self-reported and vary by app.
- 42 of 185
- founder interviews where paywall or pricing drove growth (median $41K/month where revenue was said)
- 7 days
- trial length that beat 3 days for PushScroll and, internationally, for Pingo
- +50%
- conversion lift Nicole reported for a hard paywall straight after onboarding vs letting users explore first
- +48%
- revenue Sunflower said it gained from a trial-reminder screen before the paywall
How long should a free trial be: 3 days or 7 days?
Seven days won in the two public A/B tests in the library, but the evidence is thin. PushScroll's founders had a "really bad conversion rate," ran a 3-day vs 7-day trial test, and said the 7-day trial made the difference in how many people got into the app [1]. Pingo's Mori said a 3-day trial performs about equal in the US but worse internationally, because people want 7 days; he put the international 7-day lift at about 2% [2]. Pingo also shows a 7-day trial to users who abandon a 3-day offer, and plans to test 10- or 14-day trials, while wondering how those users convert after the trial [2].
Other founders use 3 days. Superwall's paywall designer Jonathan Parra described highlighting a 3-day trial on a weekly plan, and said about half of subscribers cancel after subscribing [3]. Rizz defaults to a 7-day trial at $9.99/week, with a no-trial toggle at $6.99/week [4]. The library has no controlled test beyond 7 days, so treat 7 as a starting point to test.
Is a hard paywall better than freemium?
For the founders who tested it, a hard paywall (trial or pay before any use) usually converted better, with exceptions. Nicole of Glam Up and Sprout ran a test of a complete hard paywall after onboarding and saw a 50% conversion increase over letting users try a feature first [5]. Stronger moved from freemium to making everyone start a trial, and said revenue rose about 25% overnight with few complaints [6]. Puff Count's Stephen Cravata said a hard, unskippable paywall pushed conversion to 20-25% [7]. App Masters' Steve Young said a hard paywall doubled the trial start rate from 6% to 13% with trial-to-paid flat [8].
The counterexamples:
- Lovely (Jack Friks) went back to freemium after bad reviews of the hard paywall [9].
- Celebs (Social studio) removed its onboarding paywall and gained revenue, because later placements converted well [10].
- Halo AI combines a skippable onboarding paywall with a hard paywall at the first edit; it said this beats a single hard paywall [11].
- Stronger noted that a paid app loses its viral loop [6].
Should you sell weekly or annual plans?
It depends on the app: weekly won for several consumer apps, while annual-default won for founders who needed cash upfront. Glam Up's Nicole sold only weekly and monthly, and raising weekly from $3 to $9 increased conversion; a +20% price rise on her newer app also raised it [5]. Halo AI's standard setup is weekly ($8.99 with a trial) [11]. PushScroll found weekly beat monthly, but only by a little extra ARPU [1].
On the annual side:
- Sunflower went monthly only, then added annual, then made annual the default, and captured more annual revenue at each step [12].
- Sway AI defaulted to annual to collect cash for ad spend; weekly plans grew more slowly [13].
- Pingo pushes a $99.99 annual plan shown as a weekly equivalent, hides monthly behind "view all plans," and gives no trial on monthly [2].
Parra said weekly and monthly have become more popular for social and multiplayer apps, and that annual + weekly sometimes beats annual + monthly [3]. Nicole skipped annual for a job-search app because nobody needs a year to find a job [5].
What trial and paywall tweaks lifted conversion most?
The biggest reported lifts came from reducing fear of the trial, not from changing its length. Sunflower added a screen before the paywall saying it will remind users before the trial ends. Kobe Conrad said this gave +48% revenue, +70% annual revenue, +46% trial conversions and +30% trial starts [12]. Vahe Baghdasaryan, who runs a paywall optimization service, said a 3-screen paywall (try free, reminder, timeline) typically gives a 20-40% uplift over a single screen [4]. He also advises mentioning the free trial 5-7 times and labeling the annual card "try it free" [4].
Other results:
- Claim hit a 34% trial start rate and 50-55% trial-to-paid, then removed the trial and earned more revenue. Vahe says that needs high volume, a strong trial start and a high-intent problem [4].
- On abandonment paywalls, discount only the annual plan; he said discounting weekly or monthly didn't work in dozens of tests [4].
- Coconote offers free-trial users another 7 days in the cancel flow, which stopped 27% of cancellations [14].
Where should the paywall go in onboarding?
Most founders put it at the end of a personalized onboarding, and several said longer onboarding raised conversion. Coconote doubled onboarding from about 5 to 13 screens and said trial conversion rose 16% [14]. Copia added social proof, a video and tap-through questions; paywall view rate rose from 50-60% to 85-90% and conversion roughly doubled [15]. Clear 30 replaced the paywall's feature carousel with a day-by-day 7-day program and went from 20% to 30% conversion [16]. PushScroll rebuilt onboarding around a value-equation framework, and its paywall conversion rose from 7% to 16% [1].
Sway AI found that letting users upload screenshots before the paywall risked losing them; simple onboarding then paywall did better [13]. Beyond onboarding, Vahe lists a paywall at session start (at least once a day), after set numbers of actions or scans, and in trial-cancel and subscription-cancel flows [4]. Jungle's team argued that showing the paywall earlier reaches the roughly 50% who finish a first experience, not the 25% who reach a later step [17].
What conversion rate counts as good, and what should you test first?
Vahe said 4% install-to-paid is the minimum he'd want, and that an entertainment app went from 1.1% to about 4% with no new users [4]. Reported paywall conversion in the library runs from Bustimes' 0.5% to 8% after about 10 paywall and onboarding experiments [18] to Halo AI's 16.5% [11]. These are different metrics (paywall view vs install), so don't compare them directly.
On order of testing, Parra starts with a simple baseline, runs design and packaging tests before price tests, and says price tests get messy because of cohort management [3]. Peter of Stronger said to test macro changes (how hard, how often, what is gated) before copy, since small copy tests need months of volume [6]. PushScroll treats ARPU as its north star [1].
| Founder / app | What they tested | Reported result |
|---|---|---|
| PushScroll | 3-day vs 7-day trial | 7 days "made the difference" in how many got into the app [1] |
| Pingo | 3-day vs 7-day trial | 3-day about equal in US, worse internationally; about 2% lift for 7 days [2] |
| Glam Up / Sprout | Hard paywall after onboarding vs explore first | +50% conversion [5] |
| Stronger | Freemium to trial-first hard paywall | About +25% revenue [6] |
| App Masters | Hard paywall | Trial start 6% to 13%, trial-to-paid flat [8] |
| Sunflower | Trial-reminder screen | +48% revenue, +46% trial conversions [12] |
| Coconote | Onboarding 5 to 13 screens | +16% trial conversion [14] |
| Claim (via Vahe) | Removed free trial | More revenue after 34% trial starts and 50-55% trial-to-paid [4] |
Frequently asked questions
Does a free trial lower conversion compared with paying upfront?
Not always. Claim had a 34% trial start rate and 50-55% trial-to-paid but earned more after dropping the trial, per Vahe Baghdasaryan [4]. He says that works only with high volume and a high-intent problem [4]. Rizz instead offers a no-trial option at a lower weekly price [4].
Should I show the annual plan or the weekly plan by default?
Founders who wanted upfront cash defaulted to annual, as Sunflower and Sway AI did [12] [13]. Weekly-first apps like Halo AI and Glam Up also worked [11]. Test annual + weekly against annual + monthly, as Parra suggests [3].
Does a hard paywall hurt reviews?
Sometimes. Jack Friks switched Lovely to freemium after bad reviews of its hard paywall [9]. Stronger's founders said they had few complaints after going trial-first [6]. Steve Young said to expect some one-star reviews [8].
Sources
Founder interviews on YouTube; numbers are as the founders said them. How this research is done.
- [1]Our App Makes $30K/Month Profit Using This Simple Strategy (copy us) · PushScroll (Superwall)
- [2]How I built a $500k/mo ai app (so you can just copy me) · Pingo (Superwall)
- [3]I made 4,000 app paywalls and learned this · Superwall (Superwall)
- [4]Our No-Code Apps Made $50M Using THIS Paywall (Superwall)
- [5]she makes $400k/mo from ai apps (while in college) · Glam Up, Sprout (formerly Prep AI) (Superwall)
- [6]My App Makes $600k/Year Using This Easy Growth Strategy (copy me) · Stronger (Superwall)
- [7]How I Built It: $40K/Month iPhone App · Puff Count (Starter Story)
- [8]Beginners Guide to Buying & Selling Apps 2026 · App Masters (app acquisitions portfolio) (Superwall)
- [9]He Went From Struggling to Make $1 From His App to Turning Down $1M · Curiosity Quench, PostBridge, Doof, Lovely (Superwall)
- [10]$16M AI CEOs: How to Make $1.3M/Month With AI Apps · Social (app studio): Font Maker, Lipstick, Celebs, Autopaste, Hotspots, WiseSam.ai (Superwall)
- [11]0 to $300k/mo in 45 days with my ai app (just copy me) · Halo AI (Superwall)
- [12]I speedran an ai app from 0 to $1m/yr in 10 months · Sunflower (Superwall)
- [13]I speedran a new app from 0 to $2m/year in 6 months (here's how) · Sway AI, Cal AI (Superwall)
- [14]COPY this ai app's $6.7m/yr marketing strategy · Coconote (Superwall)
- [15]Copy This Growth Strategy, It'll Blow Up Your App · Copia (Superwall)
- [16]I Built a $30k/mo App That Helps You Quit Smoking · Clear 30 (Superwall)
- [17]this ai study app makes $100k/mo. here's how (no-code) · Jungle (Superwall)
- [18]How I built a $30K/month mobile app · Bustimes (Bomego) (Starter Story)