App Store Conversion Rate by Business Model: Free vs Paid

App Store conversion rate splits sharply by business model. A free app's install is one tap, a paid app's install is a purchase, and a subscription app's real conversion happens after the download, not at it. Apple buckets your peer benchmark by business model for exactly this reason: the funnels are not comparable [1].

So when your rate sits below the median, the problem is often that you're reading your number against the wrong bucket, not that your screenshots are weak. The three models convert on structurally different funnels, and a single blended benchmark blurs all three into one misleading average.

TL;DR:

• Apple groups peer benchmarks by category, business model, and download volume, so free, paid, and subscription apps are each compared only to their own kind [1][3].
• A paid app's page-view-to-install is a purchase decision, so it runs structurally lower than a free app's install rate. Lower is not worse, it's a different funnel.
• Freemium and subscription apps post high install conversion, but the number that pays rent is download-to-paid, which Apple now benchmarks at Day 35 [1].
• Free-trial-to-paid swings from 2.6% on a soft freemium wall to 48.8% on a credit-card trial [2]. The paywall model moves that number more than the screenshots do.

For what counts as good by category, the companion post on App Store conversion rate benchmarks covers the category dimension, and the by-country breakdown covers territory. This one is about the business model dimension, the axis most benchmark tables name in a single line and never break down.

Table of contents

• Does App Store conversion rate change by business model?
• What are the App Store business model buckets?
• Do paid apps convert lower than free apps?
• Why do freemium apps show high install conversion but low revenue conversion?
• What is a normal subscription app conversion rate?
• How should each business model design its first screenshots?
• How do you read your number against the right peer bucket?
• Reading conversion by business model

Does App Store conversion rate change by business model?

Yes, and business model is one of the three axes Apple uses to build your benchmark. Peer groups are formed by App Store category, business model, and download volume [1][3]. Apple separates business models because a free download, a paid download, and a subscription signup are different decisions with different friction, so their conversion rates don't sit on the same scale.

That design choice is the tell. If Apple thought a free game and a paid utility should be judged by the same conversion number, it wouldn't split them into separate peer groups. The blended "average app converts at X%" figure ignores the split entirely, which is why it's the least useful number you can hold your own listing to. Your business model sets the shape of your funnel before any screenshot loads.

What are the App Store business model buckets?

Apple sorts apps into five business model buckets for benchmarking: free, freemium, paid, paymium, and subscription [3]. Each carries a different install funnel, and Apple compares you only against apps in the same bucket.

• Free: no download cost, monetized elsewhere (ads, brand, data). The install is a near-zero-cost tap, so install conversion runs high.
• Freemium: free to download with in-app purchases. High install conversion, but revenue depends on a second funnel most listings never optimize for.
• Paid and paymium: an up-front price to download. The install itself is a purchase, so install conversion runs structurally lower.
• Subscription: earns mainly from auto-renewing plans. Install is usually free, and the real conversion is trial or download to paid.

Apple also refines each group by download volume into low, medium, and high tiers [3], so a 500-download-a-week app isn't benchmarked against one doing 500,000. The bucket you land in is category, then business model, then volume tier.

The takeaway from the table: two of these models are judged on the install, and two are judged one funnel further down. Comparing across those columns is where most conversion misreads start.

Do paid apps convert lower than free apps?

Usually yes, and that's expected, not a defect. A paid app's page-view-to-install is a purchase: the visitor weighs a price before tapping. A free app's install is a near-zero-cost tap. The paid funnel is structurally lower, so comparing a paid app's install rate to a free app's tells you nothing about either listing.

Paid-up-front apps are a small slice of the store now, which is part of why generic benchmarks read them so badly. Most published averages are dominated by free and freemium apps, so a paid app checking itself against "the ~25% cross-industry average" that the category benchmarks post unpacks will always look like it's failing. It isn't. The lower number is the price filter doing its job.

On a paid listing, the screenshots have a narrower assignment: justify the price before the tap. The visitor has already accepted that this app costs money, so the frames need to make the value obvious enough to clear that bar. That's a different creative goal than a free app chasing raw install volume, and it's why the two shouldn't share a benchmark.

Why do freemium apps show high install conversion but low revenue conversion?

Because the install is free, so almost nothing filters it. A freemium app can post a high page-view-to-install rate and still make no money, because the funnel that pays rent is the next one: install to paid. Apple now benchmarks that stage directly as Day 35 Download to Paid Conversion, the share of first-time downloads or redownloads that make an in-app purchase within 35 days [1].

This is where a healthy-looking listing hides a real problem. Your install conversion flatters you because it counts a free action, while the paid step quietly lags. Adapty's app-panel data puts trial-to-paid on a soft freemium wall at just 2.6% [2], so a freemium app converting 40% of page views to installs might convert only low single digits of those installs to revenue. The install rate and the money rate are two different measurements, and only one of them pays your bills.

If your install conversion looks fine but revenue doesn't follow, the leak is downstream of the screenshots. Work out which funnel stage is actually leaking before you rebuild frames that may already be doing their job.

What is a normal subscription app conversion rate?

There are two numbers for a subscription app, and they sit far apart. Install conversion is high because the download is free. The number that matters is trial-to-paid, and it swings enormously by how the paywall is built: roughly 18.2% on an opt-in trial with no card up front, 48.8% on an opt-out trial that requires a card, and 2.6% on a soft freemium wall [2].

The paywall model moves that number far more than the screenshots do, which is why a subscription app's benchmark has to account for it. By category, trial-to-paid ranges from about 1-2% for games to nearly 49% for travel [2], and over 80% of trial starts happen on the day of install itself [2]. Apple's Day 35 download-to-paid benchmark is the peer-grouped version of this same reality, letting you compare your paid-conversion against apps that monetize the way you do [1]. So a subscription app's screenshots don't close the sale, the paywall does. The frames pre-sell the value the wall will ask for, and the trial mechanics decide how many say yes. The full frame breakdown lives in the subscription screenshot pre-sell guide.

How should each business model design its first screenshots?

The job of the first frame changes by model, even though the layout vocabulary is the same. Match the frame's promise to the funnel it's actually feeding, and the number that matters moves. Get the promise wrong for the model and you optimize a metric that doesn't pay.

• Paid: justify the price before the tap. Lead with the clearest proof of value, since the visitor is making a purchase decision on the frame.
• Free: maximize the hook to win raw install volume. The install is cheap, so the frame competes purely on attention.
• Freemium and subscription: pre-sell the value the paywall will later charge for. If the frames over-promise relative to the wall, install conversion rises while paid conversion falls.

Every model still lives or dies on the first three screenshots, the ones Apple shows in search before anyone scrolls. What changes is the message they carry, not their importance.

How do you read your number against the right peer bucket?

Use App Store Connect's peer benchmark, which already groups you by business model, and stop comparing yourself to a blended vendor chart [1]. Apple shows your conversion rate against your peer group's 25th, 50th, and 75th percentile values, bucketed by category, business model, and download volume, with differential privacy so thin peer groups stay anonymous [1]. Apple's own conversion rate is total downloads and pre-orders divided by unique device impressions [1], a different denominator than the page-view rates most vendor charts quote.

The most common misread is checking a subscription app's high install rate against a paid app's low one and concluding one of them is broken. Neither is: they're different funnels. Before you trust any public table, cross-check which figures are current and which are recycled in the conversion benchmark reality-check, then hold your own number to your peer bucket, not a headline average.

Reading conversion by business model

Your business model sets the shape of your funnel before a single screenshot loads. A paid app's install is a purchase and runs low by design, a freemium app's install is free and hides its real number one funnel down, and a subscription app's screenshots pre-sell a paywall that does the actual converting [1][2]. Read each against apps that monetize the same way, which is exactly what Apple's peer benchmark already does for you.

Once you know which funnel your model runs on, the screenshot work gets sharper: a paid listing argues the price, a free listing chases the hook, a subscription listing sets up the wall. When you want to try a few frame-one directions before committing, you can describe the angle and iterate on options in the screenshot builder until one clearly fits your model, then ship it and watch the right percentile move.