Funnels
5 min read
Why aren't installs turning into trials?
You're paying for installs but barely anyone starts a trial. The gap between installed and trial started is not one step. It is five.
The gap you can't see
You've got installs coming in. Your ad spend is running. The App Store page looks fine. But when you check RevenueCat, trials are barely moving. The install-to-trial rate is sitting at 2-3% and you have no idea why.
The problem is that install to trial looks like one step. It is not. Between installing your app and starting a trial, a user passes through five or six screens, makes three or four decisions, and encounters at least one moment where they almost leave. RevenueCat tells you someone started a trial. It does not tell you what happened to the other 97% who did not.
Install -> App Open -> Welcome -> Signup -> Profile -> Home -> Paywall -> TrialYour install-to-trial rate is the product of every conversion rate between those steps. The real question is which step is weakest.
Build the funnel
Go to Conversions > Funnels and click Create funnel. Start with the full journey. You can always simplify later, but you cannot diagnose what you do not measure.
| Step | Type | Value |
|---|---|---|
| 1 | Screen | Welcome |
| 2 | Screen | Create Account |
| 3 | Goal | account_created |
| 4 | Screen | Profile Setup |
| 5 | Goal | profile_completed |
| 6 | Screen | Paywall |
| 7 | Transaction | Trial Started |
Set the conversion window to 14 days (or match your trial length). Screen views are tracked automatically by the mobile SDK — no code needed. Goal events require one line of code each (see the custom events docs). The Trial Started step comes from RevenueCat automatically — no code for that either.
Read the funnel
The funnel renders as a bar chart. Each bar is shorter than the last. The drop-off badges between them tell you where the biggest leak is.
Here's what different problems look like:
Problem 1: users never get past the welcome screen
Welcome Create Account account_created Profile Paywall Trial
5,000 800 (16%) 640 520 480 160
↓84%84% of users see the welcome screen and leave. They opened the app and immediately decided it wasn't for them. Possible reasons: the welcome screen doesn't communicate value quickly enough, the app takes too long to load, the first screen is a wall of text nobody reads, or the "get started" button isn't obvious.
This is the most common problem and the highest-leverage fix. Everything downstream is irrelevant if users bounce on the first screen.
Problem 2: users see the signup form but don't complete it
Welcome Create Account account_created Profile Paywall Trial
5,000 4,200 (84%) 1,680 (40%) 1,500 1,380 460
↓60%60% of users who see the signup form abandon it. The form itself is the friction — too many fields, confusing validation, no social login option, or the value proposition isn't strong enough to justify creating an account at this stage.
This is the second most common problem. Consider delaying account creation until after the user has seen some value, or offering a "try without signing up" option.
Problem 3: users complete onboarding but never see the paywall
Welcome Create Account account_created Profile Paywall Trial
5,000 4,200 3,360 3,100 620 (20%) 200
↓80%Users cruise through onboarding fine, but only 20% ever reach the paywall. The paywall is buried — triggered by an action most users don't take, hidden behind a feature they haven't discovered, or placed on a screen they have no reason to visit.
This is the "are users seeing your paywall?" problem. See the paywall guide for a deeper dive.
Problem 4: users see the paywall but don't start a trial
Welcome Create Account account_created Profile Paywall Trial
5,000 4,200 3,360 3,100 2,800 140 (5%)
↓95%Users reach the paywall at decent rates but almost nobody converts. The paywall itself isn't working — pricing, copy, trial terms, or design. This is the one scenario where optimising the paywall actually makes sense.
Problem 5: the whole funnel leaks evenly
Welcome Create Account account_created Profile Paywall Trial
5,000 3,500 (70%) 2,800 (80%) 2,240 (80%) 1,790 (80%) 360 (20%)
↓30% ↓20% ↓20% ↓20% ↓80%No single catastrophic drop-off, but every step loses 20-30%. This is the hardest pattern to fix because there's no single culprit. The flow is just generally high-friction — too many steps, too much effort required before any value is delivered.
Consider cutting steps entirely. Does the user really need to complete a profile before seeing the paywall? Can you show value before requiring signup? The fewer steps between install and trial, the higher the conversion rate.
Segment by source
The aggregate funnel hides the real story. Filter by traffic source and you will often see completely different patterns:
Organic (App Store) users:
Welcome → Create Account → Paywall → Trial
1,800 1,500 (83%) 1,200 (80%) 380 (32%)Organic users move through the flow efficiently. They searched for your app, they have intent, they convert.
Meta Ads users:
Welcome → Create Account → Paywall → Trial
2,000 600 (30%) 180 (30%) 20 (11%)Meta Ads users bounce hard at every step. They probably clicked a flashy ad without much intent. The install was cheap but the user quality is low.
Referral users:
Welcome → Create Account → Paywall → Trial
800 720 (90%) 640 (89%) 260 (41%)Referral users convert at nearly 3× the rate of organic. They came with a recommendation from someone they trust. Every step has minimal drop-off.
Three completely different stories, invisible in the aggregate. The "install-to-trial rate" is a blended average of channels with wildly different conversion rates. Improving it means improving the worst channels — or cutting them and spending more on the best ones.
To filter: use the filter bar at the top of Funnels. Select Source, Medium, or Campaign. The entire funnel recalculates for that segment. Use Traffic Sources to compare the same channels outside the funnel view.
The decision framework
Once you can see the funnel by source, you're making one of three decisions:
Fix the flow. If every source has the same drop-off at the same step, the step itself is broken. Fix it for everyone.
Fix the targeting. If organic users convert at 30% but paid users convert at 3%, the ads are attracting the wrong people. The flow works — the audience doesn't.
Cut the channel. If a source has a 1% install-to-trial rate and you're paying $3 per install, each trial costs $300. That might not be fixable. Reallocate the spend to the channel with a 30% conversion rate where each trial costs $10.
What good looks like
Install-to-trial benchmarks vary by app category, price point, and how aggressive your paywall is. Treat these as a rough baseline, not hard rules.
| Metric | Below average | Average | Good | Great |
|---|---|---|---|---|
| Install -> Trial | <3% | 3-5% | 5-10% | >10% |
| Welcome -> Signup | <50% | 50-70% | 70-85% | >85% |
| Signup -> Profile complete | <60% | 60-75% | 75-90% | >90% |
| Paywall -> Trial | <5% | 5-15% | 15-30% | >30% |
A meditation app with a $70/year subscription will usually have a lower paywall-to-trial rate than a photo editor with a $3/month plan. But if you are below average on any step, that is where to focus.
Try it in your data
Build your install-to-trial funnel in Conversions and check where the biggest drop-off is. Then filter by source to see if it is a flow problem or a targeting problem.
Go to Conversions ->Related guides
Funnels
6 min read
Are users seeing your paywall?
You're getting installs but no trials. The paywall might be fine — users might never reach it. Here's how to find out in 5 minutes.
Funnels
5 min read
Find the screen that kills your onboarding
Most onboarding funnels leak at one step. Set up screen-level tracking and you'll see exactly where — in about 5 minutes.