How Do Apps Make Money? The 7 Models, With Real Numbers

Apps make money in seven ways: subscriptions, usage credits, one-time unlocks, transaction fees, advertising, affiliate commissions, and paid downloads — and for a solo founder, recurring subscriptions billed on the web keep by far the most of every dollar.

If you’re planning your first app, the model you pick matters more than the features you build. It decides how much revenue survives the fees, how many customers you need to make the thing worth running, and whether you’re building a business or a hobby. Here’s how each model actually pays, with the current numbers.

The seven models at a glance

Model	How it pays	Good fit for a first solo app?
Subscriptions	Recurring monthly or annual fee	Yes — the default for most tools
Usage credits	Users buy a balance and spend it	Yes — natural fit for AI features
One-time unlock	Single payment for the “pro” version	Sometimes — no recurring revenue
Transaction fees	A cut of each booking or sale	Yes, if you’re building a marketplace
Advertising	Paid per thousand ad impressions	Rarely — needs huge traffic
Affiliate commissions	Cut of purchases you refer	As a supplement, not a plan
Paid download	Users pay before they get in	Almost never anymore

1. Subscriptions — the default for most apps

A subscription charges a recurring fee for continued access, and it’s the model behind most software businesses you’d want to own. The reason is arithmetic: every month starts with the revenue you already earned, so growth compounds instead of resetting.

It works when your app solves a problem that recurs. A scheduling tool, a client portal, an invoicing app — people need those every week, so they’ll pay every month. It struggles when the job is one-and-done, because customers cancel the moment the task is finished.

The practical benchmark: at $20/month, 100 customers is $2,000/month. That number is reachable for a focused tool serving a specific trade, and it’s a real income for one person with no employees.

2. Usage credits — the model built for AI apps

Credits let users buy a balance and spend it per action, which matches how AI apps actually cost you money. Every image generated or document analyzed calls a model that charges you. Flat-rate pricing on top of variable costs is how founders end up with a heavy user who costs more than they pay.

Credits fix that by tying what you charge to what you spend. Most AI tools now run a hybrid: a monthly plan that includes a credit allowance, with top-ups when someone runs out. That’s exactly how the AI app builders you’d use to build the app price themselves.

Watch for: users hate watching a meter. Set the included allowance high enough that normal use never hits the wall — the credits should only bite on genuinely heavy usage.

3. One-time unlock — simple, but it starts over every month

A one-time payment unlocks the full version forever, and it’s the easiest sell you’ll ever make because there’s no ongoing commitment. Buyers who won’t sign up for a subscription will often pay once.

The catch is structural. Every month begins at zero, so you have to find brand-new customers to earn the same money again. It fits small utilities and niche tools with low running costs. It’s a poor fit for anything with server or AI costs, because you collect once and pay to serve that customer indefinitely.

4. Transaction fees — you get paid when your users get paid

Take a percentage of every booking, sale, or payment that flows through your app. A booking tool for tutors that takes 5% of each lesson earns more as its users earn more, without ever raising a price.

This is the strongest model when your app sits between two sides of a transaction, and it’s the easiest to sell — nobody pays until they’ve made money. It needs real volume before the percentages add up, and you’ll need to handle payouts to your users, which is more moving parts than a simple subscription.

5. Advertising — the model that needs a crowd

Ads pay you per thousand impressions, which means the model only works at a scale most first apps never reach. Rewarded video — where a user opts in to watch an ad for something in return — is the best-paying common format, usually a few dollars per thousand impressions, though rates swing hard by country, category, and season. Banners earn a fraction of that.

Run the math before you plan on it. At a $5 effective rate per thousand impressions, a thousand daily users each seeing three ads earns roughly $15 a day. The same thousand users converting at 3% to a $20/month plan earns $600 a month. Ads are a volume business, and volume is the hardest thing for a new app to get.

Where ads genuinely fit: free consumer apps and games with high daily usage and users who will never pay. For a business tool, ads undercut the product’s credibility and earn less than a handful of subscribers.

6. Affiliate commissions — real, but secondary

You earn a commission when your users buy something you recommended. A meal planner earning a cut of grocery delivery orders, a golf app referring equipment — the revenue is real and requires no billing system of your own.

Treat it as a supplement. Commissions depend on someone else’s product, prices, and program terms, all of which can change without warning. Build a business on a model you control and let affiliate revenue be the bonus.

7. Paid download — mostly a relic

Charging before anyone has used the app worked when apps were scarce. Now it’s a wall in front of a product nobody has tried, in a store full of free alternatives.

The exception is a tool with an established reputation aimed at professionals who already know they need it. For a first app from an unknown founder, free-to-try is the only realistic path in.

The fee nobody budgets for

Here’s the part that surprises most first-time founders: whoever processes your payment takes a cut, and the size of that cut depends entirely on where the payment happens.

If you sell through the app stores:

• Apple takes 30% by default in the US. Its Small Business Program cuts that to 15% if you and any associated developer accounts took in under $1 million in proceeds last year, and subscription renewals drop to 15% after a customer’s first 12 months.
• Google Play restructured its US fees on June 30, 2026: a 10% service fee on your first $1 million in annual earnings and on all auto-renewing subscriptions, 20–25% on other transactions, plus a 5% billing fee if you use Google Play’s own billing.

If you sell on the web:

• Stripe charges 2.9% + $0.30 per US card transaction.
• Lemon Squeezy and similar merchant-of-record services charge around 5% + $0.50, and handle sales tax and VAT for you — worth the extra percentage points when you’re selling internationally and don’t want to manage tax registration.

On a $20/month subscription, that’s the difference between keeping $19.12 and keeping $14.00. Across 100 customers for a year, it’s $22,944 versus $16,800 — the same app, the same customers, $6,144 apart.

Why a web app changes your math

Every tool on our list of AI app builders except one builds web apps — and a web app skips app store commissions entirely. Your customers sign up in a browser, pay through Stripe, and you keep about 96% of a $20 charge instead of 70%.

That single decision is worth more than most pricing optimizations. It’s also why we push non-technical founders toward web-first: it’s faster to ship, there’s no review process, and the money math is dramatically better. (The one thing pushing you toward the App Store is a genuine need for the phone’s camera, notifications, or offline use — we cover that call in how to build an app.)

There’s a middle path too. Since a 2025 federal court order in Epic v. Apple, iPhone apps can link out to external payment pages, so an app can sit in the App Store and still take payment on the web. Apple is still fighting it — the Supreme Court has taken up the case — so treat this as the rule today rather than a permanent one.

So how much do apps actually make?

Most of them make almost nothing, and you should plan with that number in front of you. RevenueCat’s 2024 benchmark report on subscription apps found that 17.2% reached $1,000/month, 3.5% reached $10,000/month, and the median app earned under $50/month after its first year.

Read that as a reason to be deliberate about two things:

1. Build for a specific group of people with a problem worth paying to solve. The apps at the bottom of that distribution are mostly generic products with no particular buyer. Our scorecard is the filter we use to vet an idea before writing a line of prompt.
2. Charge from the start. Ten customers paying $30/month beats ten thousand free users, and it tells you within weeks whether you have a business. Free users tell you nothing.

The honest framing: with $20–50/month in tools and a web app you built yourself, you need roughly two or three paying customers to break even. That bar is low enough that the real question is never whether the app can be profitable — it’s whether anyone wants it.

Which model should you pick?

1. A tool people use weekly: subscription. Start at $20–50/month.
2. Anything with AI features: subscription with a credit allowance.
3. A marketplace or booking tool: transaction fee, 3–10% of what flows through.
4. A small utility with near-zero running costs: one-time unlock.
5. A free consumer app with heavy daily use: advertising, once you have the traffic.

When you’re genuinely unsure, choose the subscription. It’s the most forgiving model, the easiest to change later, and the only one on this list that pays you for work you did last year.

FAQ

How do free apps make money? Through everything except the download. Free apps earn from ads, from upgrades to a paid tier, from in-app purchases and credits, or from commissions on things users buy. “Free” is a way to get people in the door, and the money comes after.

How much does an app make per download? There’s no useful per-download number, because most downloads earn nothing. What matters is the percentage of users who pay and what they pay. A hundred downloads with three people on a $30/month plan beats ten thousand downloads with nobody paying.

Can one person make real money from an app? Yes, at a modest scale. A hundred customers at $20/month is $24,000 a year from a web app one person can run — but reaching that takes marketing work, not just building. The build is the part AI made easy; distribution is still the job.

Do I need to be in the App Store to make money? No, and for most first apps you shouldn’t be. A web app charges through Stripe with no store commission, no review process, and no annual developer fee. Go native when your product genuinely needs the phone’s hardware.

Should I charge before the app is finished? Yes — take pre-orders or sell an early-access plan. Money is the only validation that counts, and collecting it early tells you whether to keep building. We walk through this in the 48-hour validation sprint.

The bottom line

Pick the subscription unless you have a specific reason not to, price it at $20–50/month, and sell it on the web so you keep about 96% of it instead of 70%. Then spend your energy on the part that actually decides the outcome: finding people who have the problem and telling them you solved it. That’s the whole game, and it’s what Software Secrets 2.0 covers start to finish. The book is free.