When Building Is Free, Who Gets to Charge?

Apple's App Store had 557,000 new submissions in 2025. That's up 24% from the year before, according to Appfigures. A lot of those aren't from developers having creative breakthroughs. They're the same ideas, cloned at the cost of a weekend.

Building an app used to mean $50,000 and several months. With AI coding tools now baked directly into Xcode, it's starting to mean a few days and an API bill. That gap matters, and it changes everything downstream.

The pricing logic doesn't hold anymore

Subscription pricing on software always had a loose justification: you're not paying for the binary, you're paying for ongoing development, support, and infrastructure. For local apps (PDF editors, file converters, text tools that never phone home), that justification was always shaky. Now it collapses.

If I can clone your $10/month local PDF editor over a weekend, someone will. Then they'll price it at $5. Then someone else makes it free to grow their user base. Then the original developer has a churn problem and a positioning problem at the same time.

The race to zero on local utility apps isn't a theory. It's already underway.

Server-dependent apps buy more time, not safety

Apps that sync data, run AI features, or depend on backend infrastructure still have real costs to cover, so subscription pricing holds up longer. But the pressure follows. If the only reason to maintain a subscription is server costs, users will eventually find a self-hosted version or a cheaper clone that runs its own servers.

The businesses that keep pricing power aren't the ones with the most features. They're the ones with moats that take longer than a weekend to replicate.

Network effects are the strongest version of this. A tool that gets more useful as more people use it doesn't lose value when you clone the code. Notion is harder to commoditize than a PDF editor because a significant chunk of its value sits in templates, integrations, and shared workspaces that other users created. You can copy the product. You can't copy the network.

Data flywheels work similarly. If your product improves as users interact with it (smarter defaults, better recommendations, learned preferences), copying the codebase doesn't copy three years of behavioral signal. The longer users have been generating data, the wider the gap gets.

Integration surface area creates switching costs that survive a price war. When your API powers other tools in someone's stack, those tools break if you leave. A clone has to replicate not just the product but the entire dependency graph around it.

Human accountability is the underrated one. Anything where someone needs to sign off, audit, or explain a decision to a regulator doesn't get commoditized by AI. People in regulated industries want a vendor they can call. That's worth paying for.

Apple's move here is worth paying attention to

I keep coming back to the fact that Apple didn't tighten App Store review when the submission flood started. They accelerated it by integrating AI coding tools directly into Xcode.

That's not an accident. Their revenue scales with app volume even if per-app revenue shrinks. They're the platform. Commoditization fills niches that were previously too small to build for, which is good for them. They win in a world where building costs nothing.

For developers, this split is uncomfortable. The platform wins. Apps with real moats win. Everything in the middle gets squeezed: the subscription local utility with no network, no data flywheel, and no integration story.

What I'd build today

Not a local utility with a subscription attached. That model has maybe 12 to 18 months of comfortable margins left, and I wouldn't want to be defending it when the floor drops.

I'd build things that become more valuable over time. Not because of shipping features, but because of accumulated state. User-generated content. Learned behavior. Network density. Integration depth that grows with each new connection.

Features get cloned. State doesn't.

The developers who internalize that distinction early will be building sustainable businesses. Everyone else will be racing to the bottom and wondering why the economics stopped working.

---

About the Author: Muhammad Khan is a Principal Full Stack Engineer with 9+ years of experience building scalable systems for millions of users.