I get asked this once a year, and the 2026 version has more people tied up in knots than usual. My take isn't the comforting kind: it was never a "should you" question. It's "what kind of app are you actually building."
Conclusion first: if your app is for internal enterprise use, still in beta, or fundamentally a utility — nobody is going to distribute it for you, because app stores simply weren't designed for those cases. But if you're building a consumer app that lives on store traffic, listing honestly is still the highest-ROI path, and self-hosting is at best a supplement.
The math isn't complicated. On the store side: the 30% cut is real money, review unpredictability is real, and rejection reasons can be baffling — I get the complaints. But the store solves one genuinely hard problem for you: where users come from. An indie dev who hosts the app on their own server saves 30% and then has to single-handedly solve acquisition, trust, payments, and update delivery. For most small teams that trade loses money.
Flip it around, and the cases where self-hosting clearly wins are specific: internal apps for employees that could never list; beta builds that need fast iteration and ad-hoc invites; utility apps (system enhancements, accessibility tools) that fail review in most stores; and staged rollouts, A/B tests, per-channel packaging stats. In these scenarios the store isn't "expensive" — it's unusable.
So my answer to "do indie devs still need self-hosted distribution" is: stop treating it as either/or, split it by scenario. If you need to reach strangers and live on store traffic, spend your energy on the store. If you face something unlistable, or your audience is clearly controllable, build your own channel — it'll go more smoothly. The worst move is to hear "the store cut is brutal, going solo is cheaper," go all-in, and then spend everything you saved trying to buy acquisition and trust back.