Skip to content
Expert Guide Series

Free vs Paid Apps

A client came to us with an app full of backing tracks for musicians, and a price already fixed in their head. That price came from what the recordings had cost to make in the studio. We recommended going low to get mass adoption, because the market was niche and split between working professionals and people playing at home for fun. The client pushed for the high price instead. Sales came in low, as we expected.

The price you set decides who turns up, and that changes everything you design afterwards.

That project is the cleanest example we have of how free vs paid apps goes wrong. The money question gets settled first, usually from the inside, and then the product gets built around an answer nobody tested.

So this piece works through the decision properly. What the four real models are, how to match one to your market, where the paywall goes, how to run free and paid versions of the same app, what the platforms take, and what the numbers look like when you've chosen badly. Most of it comes from products we've worked on, including the ones where our advice was ignored and we got to watch what happened next.

The Four Models You Are Actually Choosing Between

People talk about free or paid app as if it's a switch with two positions. There are four real options, and each one asks something different of the product.

ModelWhat the user givesWhen they decideWhere it breaks
Paid upfrontOne payment before opening itIn the store, with no experience of the productThe listing has to carry all the persuasion
FreemiumNothing, then money for moreAfter they've used it for a whileFree tier too generous, or too thin to judge
SubscriptionA fee every month or yearUsually in the first weekValue has to keep arriving, forever
Free with ads or in-app purchasesAttention, or small amounts oftenOver and overThe ads undercut how the product feels

Paid upfront is the hardest of the four, and it's getting harder. As of May 2025, Statista, 2025 put 95.41% of iOS apps as free to install. People have learned that trying costs nothing, so asking for money before they've touched anything works against a habit built over years.

Subscription and freemium often get treated as the same thing. They're different. Freemium is a permanent free product with a paid layer on top. Subscription is one product with a meter running.

Matching the Model to Your Market and Audience

The model has to fit the size of your market and the way your audience already buys things. On the backing track app, the market was niche and it had two very different groups inside it. A session musician with a gig on Friday thinks about cost differently to someone learning at home. One price had to serve both, which is exactly why we argued for a low one and for volume.

Free vs paid apps in a niche market

Small markets punish high prices twice. You lose the casual end completely, and the professional end is too small to make the numbers work on its own. Low pricing with volume gives you a base of users to learn from, which is worth more in year one than the revenue is.

Platform matters too. AppTweak reports that iOS users typically spend two to three times more on apps and in-app purchases than Android users, who lean towards free, ad-supported products. Treat that as a rule of thumb rather than a measurement, but if your audience sits mostly on Android, a paid upfront launch is fighting the grain.

So the useful question is how your audience already pays for things like yours. Beauty and skincare buyers are used to subscriptions. Fitness buyers are used to tiers. Follow the habit they have.

Design that understands your users

We build app experiences around real user behaviour, not assumptions. Research, psychology-driven design and technical specs that turn users into loyal advocates.

See how we work Get started

No commitment

Setting the Price: What Production Value Has to Do With It (Nothing)

The client on the music app came from the recording world, and they knew to the penny what the sessions had cost. They wanted that money back within weeks. As Simon put it, "they wanted to immediately recoup that cost within weeks kind of thing and it's just that's just not how things work."

That anchor is the most common pricing mistake we see, and it's not really a maths error. It's emotional. You've spent a year and a large budget making something good, and setting a low price feels like saying the work was cheap.

What something cost you to make has nothing to do with what a person will pay for it.

Buyers don't know your production costs and they don't care. They price against the last comparable thing they paid for, and against free alternatives that are one search away. Garena's Free Fire took serious market share by pricing in-app purchases more than 50% below comparable games, according to Liquid & Grit.

If you can't separate the two in your head, write down your price, then write down what you'd pay as a stranger with no knowledge of the build. The gap between those two numbers is the problem.

Running Free and Paid Versions of the Same App

Yes, you can run both, and there are two ways to do it. One app with a free tier and a paid tier inside it, or two separate listings in the store. The first is almost always the better choice.

Two listings split your reviews, your ranking and your install base across two products. It also forces people to download a second app to pay you, which is a strange thing to ask of someone who has just decided to give you money. One app with tiers keeps the history, the settings and the progress intact when they upgrade.

If you already run a separate "pro" listing, keep it alive for existing buyers but point new traffic at the single app. Migrating people mid-flow loses more of them than the tidy-up is worth.

The real work in running free and paid versions of the same app is deciding what the free version is for. It has three possible jobs.

  • A demonstration, where the free tier proves the product works and stops short of the main benefit
  • A habit builder, where the free tier is genuinely useful forever and paid adds depth
  • A reach play, where free carries ads and paid removes them

Pick one. Products that try all three end up with a free tier that confuses people and a paid tier nobody can describe.

Where to Put the Paywall

A paywall placed before anyone has felt the product work is a cold ask. A paywall placed after the first real moment of value is a reasonable one. Everything else is detail.

So the useful exercise is mapping what the first genuine payoff is in your product, then putting the wall just past it. In a wellness app that's the first session finished, not the sign-up. In a fashion resale app it's the first listing going live. People pay once they have something to lose.

Layering instead of blocking

There's a difference between hiding things and layering them. Progressive disclosure gives people more depth as they ask for it, so the product grows with their understanding. A hard paywall in the middle of a flow does the opposite, stopping someone mid-thought with no sense of what sits behind it.

We'd always rather show what the paid tier contains and let someone look at it than grey it out with a padlock. Seeing the thing you can't have yet creates a clear reason to upgrade. A padlock creates mild irritation and a guess.

Write the sentence a user says to themselves at the paywall. If it's "I can see exactly what I get", you're fine. If it's "what even is this", move the wall later.

Freemium Done Badly: Pay-to-Win and Other Hostage Tactics

The old pay-to-win model in gaming is the clearest version of freemium done badly. You couldn't finish a level without paying, so the payment wasn't buying value, it was buying relief from a problem the designers had built on purpose. That approach has mostly died off and been replaced with earned engagement and earned rewards, which is a much healthier way to do it.

The same tactic shows up in gentler clothing. Artificial waiting times you can skip for a fee. Daily limits set just below the point of usefulness. Data you've entered yourself that you can't export without upgrading. People recognise all of these, and the feeling they leave behind sticks to the brand.

There's a related trap worth naming. We worked on a product where gamification carried the whole engagement model, and it worked in a functional sense. Simon's reading of it was blunt, that "although the core product worked really well in a functional sense emotional, it was completely flat." Strip the points and badges out and there was nothing underneath for anyone to feel.

Gamification lands better when it rewards behaviour rather than outcomes. Showing up, checking in, doing the thing regularly. Rewarding results alone filters out everyone who isn't already winning.

What Each Model Does to Your Onboarding

Each model changes what the first two minutes have to achieve. Paid upfront buys you patience, because the person has already committed money and wants it to be worth something. Free gives you no patience at all.

The commonest self-inflicted wound in free apps is forcing registration before anything useful has happened. We see a sharp rise in uninstalls at that exact screen, and it lands inside the first minute or two, before the product has made its case.

Forcing registration in the first minute costs you installs before anyone has seen the product work.

What helps is mapping the use case rather than the screen. How did this person arrive, what was happening in their life just before, what state are they in. Someone downloading a sleep app at eleven at night is in a different frame to someone browsing it at lunch. If you only design from the splash screen forward, you've thrown away everything that explains them.

For subscription products, onboarding has one more job. It has to set the expectation of what arrives next week and the week after, because that's what the renewal is really paying for.

Being Honest About Money: Fees, Trials and Renewals

On a travel booking product, we wrapped the platform's Stripe booking fee into the headline total. The thinking was that travellers want one clean number. What we found was the opposite. People expected a platform fee as a separate line, because that's how they've learned apps and booking sites work. By hiding it inside the total, we created a worry that a surprise charge was coming later.

So we broke everything out and showed each fee plainly. Confidence went up while the amount of information on screen also went up. In Simon's words, "there was this misconception that there was going to be an extra fee added later", and showing more removed it.

The same principle runs through trials and renewals. Say the renewal date in plain words, say the amount, and make cancelling findable. A trial that quietly converts earns you one payment and loses you the person.

Transparency does have a limit worth respecting. Risks and costs land properly when they sit next to the reason the thing is worth doing. Lead with nothing but charges and conditions and people switch off before they get to the value.

Retention and Churn: What Changes After the Transaction

Money changes the relationship. A free user who stops opening the app drifts away quietly. A paying user who stops opening it becomes a cancellation, a refund request or a one-star review, and all three are visible to everyone else considering you.

So paid products have to keep proving themselves on a schedule. Monthly subscriptions get judged monthly. Annual ones get judged in one brutal moment twelve months later, usually by someone who hasn't opened the app since March. The quiet period before an annual renewal is where most of the damage happens.

Three weeks before an annual renewal, show the person what they actually did with the product that year. Sessions, saves, workouts, orders, whatever your unit of use is. Facts beat persuasion at the renewal screen.

Free tiers need a different kind of attention. The goal there is a habit strong enough that the paid tier becomes an obvious next step rather than a sales pitch. That means the free experience has to be emotionally complete on its own, with a real payoff each time, instead of feeling like a demo with the good parts removed.

Churn is also a design signal. People rarely leave over price. They leave because the thing stopped fitting into their week.

App Store Rules, Commission and Platform Constraints

Whatever model you pick, Apple and Google take a cut of in-app purchases. SplitMetrics sets out the structure, with fees sitting in a 15% to 30% band depending on your size and the type of purchase. Apple charges 15% for developers in its Small Business Programme under a million dollars a year, and 15% on subscription renewals after the first year against 30% in year one. Google takes 15% on the first million of annual earnings and 30% above it.

Build that into your pricing from the start. A nine ninety-nine subscription is not nine ninety-nine of revenue, and plenty of plans we've reviewed were drawn up as though it were.

The rules also shape what you can say inside the app. Pointing people to a cheaper option on your website has been a long-running fight between developers and platforms, and the detail keeps moving, so check the current position before you design a flow around it.

One thing stays constant. Subscriptions are managed in the user's platform account, so cancelling happens somewhere you don't control. Explain where that is and make it easy to find. Hiding it doesn't keep anyone, it just adds anger to a decision already made.

The Numbers That Tell You the Model Is Wrong

Most teams watch downloads, which tells you about marketing. The numbers that tell you the model is wrong sit further down.

Freemium conversion is the first one. Databox puts typical free-to-paid conversion at 3% to 5%, with no sample detail given, so treat it as a rough benchmark rather than a target. If you're well under it while free usage looks healthy, the free tier is doing the whole job on its own.

When free users are happy and nobody upgrades, the paid tier has no job that anyone values.

The opposite reading matters too. High conversion with high churn a month later means the paywall is catching people before they understand what they bought.

Here's what we'd look at, in this order.

  1. How many people reach the first real moment of value, and how long it takes them
  2. What share of those people see the paywall at all
  3. Conversion from seeing the paywall to paying
  4. Thirty-day retention, split between free and paid
  5. Refund requests and the reasons given in reviews

Low numbers at step one are a product problem. Low numbers at step three with healthy numbers above it are a pricing or placement problem. The difference matters, because teams fix pricing when the product was the issue all along.

Changing Model After Launch

You can change model after launch. The thing to protect is the people who already paid, because they're watching how you treat them and so is everyone reading the reviews.

Going from paid to free is the more sensitive move. Anyone who bought the app yesterday now sees it given away, and that feels like a loss even though they've lost nothing. Grandfather them into the top tier permanently and tell them you've done it, by name, in a message rather than a release note.

Going from free to paid is slower but calmer. Keep the existing experience free for existing users, put the paid tier around new capability, and let the change arrive as an addition rather than a removal.

On the backing track app, a shift to lower introductory pricing or a volume model was open to the client at any point after launch, and our view was that both downloads and total revenue had room to climb. The high price stayed, so the product never reached what it was capable of commercially. The pricing decision outlived every other decision on that project.

So treat your model as a thing you can revisit at six months, with the numbers in front of you, rather than a promise you made to yourself before launch.

Conclusion

Free vs paid apps is a question about who you want in the product and what you want them to feel at the moment money comes up. Price filters your audience before anything else does. A high price on a niche product keeps out the people who make the thing work, and a free tier that gives away the whole job leaves nothing to sell.

Three things carry most of the weight. Price against what people pay for comparable things rather than what your build cost. Put the paywall just past the first moment the product proves itself. Show every fee, every renewal date and every amount in plain language, because hiding a charge creates a bigger fear than the charge itself.

And check the model again at six months. The decision you made before launch was a guess with no users in it. The one you make at six months has evidence behind it.

If you're weighing up a free or paid app, or your current model isn't converting the way you expected, we're happy to look at it with you. Let's talk about your app's pricing model.

Frequently Asked Questions

What are the four pricing models I should be choosing between?

They are paid upfront, freemium, subscription, and free with ads or in-app purchases. Each one asks something different of the product, from a store listing that has to do all the persuading, through to a free tier that must be generous enough to judge but thin enough to convert.

Is charging upfront still viable?

It is the hardest of the four models and it is getting harder, with Statista putting 95.41% of iOS apps as free to install as of May 2025. People have learned that trying costs nothing, so asking for money before anyone has touched the product works against a habit built over many years.

What is the difference between freemium and subscription?

Freemium is a permanent free product with a paid layer sitting on top of it, so users can stay on the free tier indefinitely. Subscription is one product with a meter running, which means value has to keep arriving month after month or people cancel.

Should I price low or high in a niche market?

Small markets punish high prices twice, because you lose the casual end completely and the professional end is rarely big enough to carry the numbers alone. Low pricing with volume gives you a base of users to learn from, and in year one that learning is usually worth more than the revenue.

Does it matter whether my audience is on iOS or Android?

Yes. AppTweak reports that iOS users typically spend two to three times more on apps and in-app purchases than Android users, who lean towards free, ad-supported products. Treat that as a rule of thumb and check it against your own audience rather than assuming it holds.

What happens if my audience is split between professionals and hobbyists?

A single price has to serve two groups who think about cost very differently, which is a strong argument for pricing low and going for volume. A session musician with a paid gig and someone playing at home for fun will not respond to the same number, so pricing for one usually means losing the other.

Why does the pricing decision need to be made before the product is built?

The price you set decides who turns up, and that shapes everything you design afterwards. When the money question is settled internally without testing, the product ends up built around an answer nobody has validated with real users.

Can I run free and paid versions of the same app?

Yes, and it is a common approach, but you need to be clear about where the paywall sits and what each version is for. The risk is a free tier that is either too generous to convert anyone or too thin for people to judge whether the paid version is worth it.