An app with fifty thousand downloads earns nothing if nobody opens it twice. That single sentence explains most of the disappointment developers feel after launch, and it is why choosing a monetization model before understanding your retention is the wrong order.

There are three ways a mobile app makes money: showing ads, selling something inside the app, or charging a recurring fee. Each one demands something different from your users, and picking the wrong one for your app is a mistake that takes months to undo.

Advertising: easy to add, hard to earn from

Ads are where almost every solo developer starts, because the barrier is so low. Add an SDK, place a banner, done. No pricing decisions, no payment handling, nothing to persuade anyone of.

The problem is the arithmetic. Ad revenue is impressions multiplied by eCPM, and impressions come from sessions rather than installs. An app with two thousand daily active users seeing three ads per session at a four dollar eCPM earns somewhere in the region of a few hundred dollars a month. That is a real number, and for many developers it is far below what they expected from an app with tens of thousands of downloads.

Ads work when you have volume and frequent sessions. Utilities people open daily, casual games, news readers. They work badly for apps used once a month, however useful that one use is.

Not all ad formats are equal

• Banners earn least per impression but are the least intrusive. Persistent, low value.
• Interstitials earn considerably more but interrupt the user. Place them at natural breaks, never mid-task.
• Rewarded video typically earns the most per impression and users opt into it willingly, because they get something in return. If your app has anything that can be unlocked, this is usually the format to prioritize.
• Native ads sit inside your content and preserve the experience, but they take design work to do well.

The most common error is raising ad frequency to raise revenue. It works for a few weeks, then retention falls, ratings drop, and total revenue ends up lower than before. You can model this with our app revenue calculator — but the honest test is whether the app survives the increase.

In-app purchases: sell something specific

IAP means selling something discrete inside the app: removing ads, unlocking a feature set, buying credits or currency.

The economics are far better per user than advertising. A single ad-removal purchase can be worth more than that user would generate in ad revenue over a year. And it does not require huge scale, which is why it often suits smaller apps better.

Conversion rates are low. A small percentage of users pay for anything, and the rest never will. That is normal and not a sign of failure. The model works because the paying minority more than covers the cost of serving everyone.

Making IAP work

• Let the free version be genuinely useful. A crippled free app gets uninstalled before anyone considers paying.
• Ask at the moment of value. Prompt after someone has used the feature successfully, not on first launch.
• Keep the offer simple. One clear unlock beats five confusing tiers.
• Price for your market. Store pricing can be set per region, and a price that works in the US may be far outside reach elsewhere.

Combining ads with an ad-removal purchase is the classic pairing and it works well: non-payers generate ad revenue, and the ads themselves are the argument for buying the upgrade.

Subscriptions: the best economics, the hardest promise

Subscriptions produce predictable, compounding revenue. A thousand subscribers at five dollars a month is five thousand dollars that arrives again next month without new acquisition. Nothing else in mobile does that.

The price is that you have to keep earning it. A subscription is a promise of ongoing value, and users cancel the moment that promise stops being kept. Churn is the metric that decides whether the model works, and it is unforgiving.

Subscriptions suit apps that deliver value repeatedly and visibly: content that updates, cloud sync across devices, ongoing tracking, anything with a service behind it. They suit one-off tools very badly. If your app solves a problem once and completely, a subscription is the wrong shape and users will say so in reviews.

Free trials and their trade-off

Trials raise conversion substantially, because people commit more readily once they have used something. They also attract users who cancel before the first payment, so the headline trial-start number is misleading.

Track trial-to-paid conversion rather than trial starts. That is the figure that tells you whether the trial length is right and whether the app delivers value fast enough within it.

Store fees you need to price around

Google takes a service fee on in-app purchases and subscriptions. The standard rate has historically been higher for larger developers and reduced for smaller ones and for subscription revenue, and the specifics have changed several times following regulatory decisions in multiple countries.

Because this area moves, check the current Play Console fee schedule before setting prices rather than relying on a figure quoted in an article. What matters for planning is that a meaningful share of every transaction goes to the platform, and your pricing has to survive that.

Advertising revenue is not subject to store fees in the same way, though the ad network takes its own share before you see anything.

The numbers behind each model

It helps to see roughly what each model produces at the same audience size, because the gap surprises people. Take an app with 5,000 daily active users.

On advertising, with two sessions a day and three ad impressions per session at a four dollar eCPM, you are generating around 30,000 impressions daily. That produces roughly a hundred and twenty dollars a day before the network’s share, so somewhere in the low thousands per month. Respectable, and entirely dependent on those users continuing to open the app.

On in-app purchases, if two percent of those users ever buy a five dollar unlock, that is a hundred purchases from the current user base. It is a one-off rather than recurring, so the revenue depends on new users arriving continuously.

On subscriptions, if three percent convert to a five dollar monthly plan, that is a hundred and fifty subscribers producing seven hundred and fifty dollars a month — and it repeats, growing as long as new subscribers outpace churn.

The subscription figure looks smallest in month one and largest by month twelve. That is the entire argument for the model, and also why it is the hardest: you have to survive the early months when it earns least.

Churn is the number that kills subscriptions

Monthly churn is the percentage of subscribers who cancel each month, and it sets a hard ceiling on how large your subscriber base can ever get. At ten percent monthly churn, the average subscriber stays ten months. At five percent, twenty months. That difference doubles the lifetime value of every single customer.

It also caps growth. If you add five hundred subscribers a month and churn ten percent, your base stabilises around five thousand and stops growing regardless of how much you spend on acquisition. Halving churn moves that ceiling to ten thousand without acquiring a single extra user.

• Cancellations cluster early. Most happen in the first month, usually because the app never delivered its promise fast enough. Onboarding is a churn problem.
• Annual plans reduce churn structurally. A yearly subscriber cannot cancel monthly. Offering a discount for annual billing usually pays for itself.
• Ask why on the way out. A one-question cancellation survey is the cheapest research you will ever run.

Common mistakes across all three

• Monetizing before retention works. The most expensive mistake, and the most common. Revenue applied to a leaking app leaks with it.
• Asking too early. A purchase prompt on first launch converts poorly and raises uninstalls. Let people experience the value first.
• Changing model repeatedly. Switching from ads to subscription to purchases within a few months confuses existing users and generates angry reviews that outlast the experiment.
• Ignoring regional pricing. A single global price is either too high for most of the world or too low for the markets that would pay more.
• Forgetting the data safety declaration. Adding an ad SDK changes what your app collects, and your Play Store declaration has to be updated to match. This is a policy issue, not a technical one.

Frequently asked questions

How many users do I need before ads are worth it?

There is no fixed threshold, but below a few thousand daily active users, ad revenue is usually too small to matter. Work it out for your own figures rather than guessing — revenue per daily user is the number that tells you, not install count.

Can I use ads and in-app purchases together?

Yes, and it is the most common combination that works. Non-paying users generate ad revenue while the ads themselves make the case for the ad-removal purchase.

Should I launch paid instead of free?

Paid-upfront apps are a difficult sell for an unknown developer, because nobody can try before buying. A free app with a paid unlock reaches far more people and lets the product do the persuading.

When should I add monetization?

Once you can see that a meaningful share of users return after a week. Before that, any revenue model is decorating a problem rather than solving one.

Choosing, in practice

• Opened daily, large audience, short sessions? Advertising, with rewarded video where possible.
• Clear premium feature people would pay once to unlock? In-app purchase, probably alongside ads.
• Ongoing value that renews, or a service behind the app? Subscription.
• Small audience, high individual value? Paid unlock or subscription. Ads will not produce meaningful revenue at low volume.

Most successful apps end up combining two. Ads plus an ad-removal purchase is the most common pairing, and a free tier with a subscription upgrade is the next.

Fix retention first

This is the part worth repeating, because it decides everything above. All three models depend on people coming back. Ads need repeat sessions, purchases need enough engagement for someone to want more, and subscriptions need continuing value.

An app with strong retention and no monetization can be monetized later. An app with monetization and weak retention cannot be rescued by switching ad networks, and developers spend months trying.

If your app is not live yet, our guide to publishing on the Play Store covers the testing requirements that come first. If it is live, spend the next month on why people leave rather than on how to charge the ones who stay.

Both models reward the same underlying thing: an app people choose to open again tomorrow. Get that right and any of the three will work. Get it wrong and none of them will.