Updated October 30, 2026
What paywall conversion is
Paywall conversion is the share of people who saw your app’s paywall and bought. It is the most direct way to know whether the pricing screen is doing its job, because it ignores everyone who never reached it.
It is different from install conversion. An app can have a good paywall conversion and a low install conversion if few people get to the paywall. That is why it pays to measure both and never mix them.
In subscription apps, the paywall is usually the decision point of the whole business: everything before it (ad, store page, onboarding) exists to bring people there, and everything after it (retention, renewals) depends on them buying there.
Why measure people, not views
The paywall usually shows up more than once for the same person: at the end of onboarding, when they tap a paid feature, when they open the app again. If every view goes into the denominator, the rate drops even though nothing changed in how people behave.
An example: 1,000 people saw the paywall, adding up to 2,500 views, and 50 bought. Dividing by views, conversion is 50 ÷ 2,500 = 2%. Dividing by people, it is 50 ÷ 1,000 = 5%. Both use the same data, but only the second answers how many people who saw the price decided to pay.
| How you count | Math | Result |
|---|---|---|
| Per view | 50 ÷ 2,500 | 2.0% |
| Per person | 50 ÷ 1,000 | 5.0% |
How to calculate it
The math uses people, not views:
Paywall conversion = exposed payers ÷ unique exposed users × 100
- Pick the period and the cohort, for example everyone who installed last week.
- Count the unique users who saw the paywall in that period.
- Count how many of those users made a confirmed purchase.
- Divide payers by exposed users and multiply by 100.
Worked examples
Three scenarios with the same 6% target. The right-hand column is what the calculator above returns as “additional payers for target”.
| Exposed users | Payers | Conversion | Payers for 6% | Missing |
|---|---|---|---|---|
| 1,000 | 50 | 5.0% | 60 | 10 |
| 2,500 | 150 | 6.0% | 150 | 0 |
| 8,000 | 320 | 4.0% | 480 | 160 |
Paywall conversion vs install conversion
Picture 10,000 installs. Of those, 4,000 people reach the paywall and 120 buy. Paywall conversion is 120 ÷ 4,000 = 3%. Install conversion is 120 ÷ 10,000 = 1.2%.
The two tell different stories. If paywall conversion is good and install conversion is low, the problem is earlier: in the onboarding, which loses people along the way. If both are low, look at the paywall and the offer first.
| Metric | Math | Result |
|---|---|---|
| Paywall conversion | 120 ÷ 4,000 | 3.0% |
| Install conversion | 120 ÷ 10,000 | 1.2% |
| Reach paywall | 4,000 ÷ 10,000 | 40% |
The whole funnel, step by step
Paywall conversion is one step of a bigger funnel. Looking at the whole funnel shows where people drop off. In the same 10,000-install example, with 6,000 people finishing onboarding and 800 starting a trial:
| Step | People | From previous step | From installs |
|---|---|---|---|
| Installed | 10,000 | — | 100% |
| Finished onboarding | 6,000 | 60% | 60% |
| Saw the paywall | 4,000 | 66.7% | 40% |
| Started a trial | 800 | 20% | 8% |
| Paid | 120 | 15% | 1.2% |
Paywalls with a free trial
When the paywall offers a trial, paywall conversion is the product of two rates: how many start the trial and how many pay when it ends. In the funnel above, 800 of 4,000 started a trial (20%) and 120 of 800 paid (15%). 20% × 15% = 3%, the same paywall conversion.
Splitting the two rates shows what to change. If few people start the trial, the problem is the offer and the paywall. If many start and few pay, the problem is the experience during the trial or the billing reminder.
Paywall conversion = trial start rate × trial-to-paid conversion
Calculate trial-to-paid conversionFrom conversion to revenue
One point of conversion turns into money fast. With 1,000 people seeing the paywall and a $29.90 monthly subscription, each conversion point means 10 more subscribers, or $299 of monthly recurring revenue for every 1,000 people.
| Conversion | Subscribers | Monthly revenue ($29.90) |
|---|---|---|
| 4% | 40 | $1,196 |
| 5% | 50 | $1,495 |
| 6% | 60 | $1,794 |
Where each number comes from
Both numbers come from your app’s events, in the subscription or analytics tool you use:
- Exposed users: the paywall view event, counted per unique user and not per view.
- Payers: purchases confirmed by the store, tied to the same users and period.
- Period: the same for both numbers. Mixing the week of who saw it with the month of who bought distorts the rate.
- Platform: measure iOS and Android separately before adding them up, because price, store and audience can differ.
How to set a target
The most useful target starts from your own baseline, not from a market average. Measure your current conversion for a few weeks to learn how much it moves on its own, and only then decide how much you want to raise it.
An example: with 4,000 people seeing the paywall each week and 3% conversion, that is 120 payers. A 3.5% target needs 140 payers, or 20 more per week. At $29.90 a month, that adds $598 of monthly recurring revenue for each week of new users. Use the calculator above with your target to see how many payers are missing.
What moves paywall conversion
No single lever works for every app. These are the things usually tested, one at a time:
- Where the paywall shows up: at the end of onboarding, after showing a result, or only when someone tries a paid feature.
- What comes before it: questions that personalize the promise and screens that show value before the price.
- Free trial: whether to have one, how long it lasts and how it is explained.
- Plans: which ones show, which one is highlighted and whether the annual price is shown per month.
- Copy and proof: the paywall headline, the list of benefits and user reviews.
How to test a paywall change
Changing the paywall without measuring is trading one opinion for another. A simple test follows these steps:
- Write the hypothesis: what changes, for whom and what result you expect.
- Change one element at a time, so you know what caused the difference.
- Calculate up front the sample size needed to detect the difference that matters.
- Run the test for full weeks, so you don’t compare weekdays with weekends.
- Decide by the criterion agreed before starting, and check the significance of the result.
How to study competitors’ paywalls
Before testing in your app, it pays to see what apps that already make money do. In Noctral’s onboarding and paywall library, each app shows up screen by screen, in the real order, with every screen tagged by stage and the app’s estimated revenue alongside. You can see where the paywall comes in, what comes before it and how the trial is offered.
See the onboarding and paywall libraryCommon mistakes
- Counting views instead of people: someone who sees the paywall five times enters the denominator five times and the rate drops for no reason.
- Counting a started trial as a purchase: trials are a different metric. For this rate, only confirmed purchases count.
- Comparing different periods: a holiday week and a regular week are not the same cohort.
- Ending the test early: a good result in the first days can vanish once the sample is complete.
- Forgetting refunds: if they matter, also track conversion net of refunds.
See it in real data
Onboarding and paywall
Screens from apps that already earn, step by step
Frequently asked questions
How do I use the paywall conversion calculator?
Measure confirmed purchases among unique users exposed to a paywall, without mixing repeated impressions with people.
Should I count repeated paywall views?
Not in the denominator of this user-based rate. Deduplicate people and count at most one converter per person.
Does a free trial count as a conversion?
Not in this rate. Only confirmed purchases count here. Trials started and trial-to-paid conversion are separate metrics, and both are worth tracking.
What is a good paywall conversion?
It depends on price, audience, whether there is a trial and where the paywall shows up. Compare your rate with itself over time and across paywall versions, more than with market averages.
How long does a paywall test need to run?
Until it reaches the sample size you calculated before starting, covering full weeks. Stopping when the result looks good raises the chance of a false positive.
Should I measure iOS and Android together?
Measure them separately first. Price, commission, payment method and audience can differ between stores, and the average of both can hide a problem in one of them.
Do I need an account and are my inputs uploaded?
No account is required. Calculations run in your browser. Entered numbers and text are not included in tool analytics events; only the tool ID, language, and action are recorded.