iOS is the most lucrative consumer app market dollar-for-user but the most opinionated about how you can monetize. This post is the practical economics specifically for iOS App Store apps in 2026. For the shared framework, see App Economics Framework.
Apple's fees
- Apple Developer Program: $99/year. Required for App Store distribution. Cheap.
- App Store standard commission: 30% on paid app sales, in-app purchases, and first-year auto-renewing subscriptions.
- Reduced commission for second-year subscribers: 15%. Reward for retention.
- Apple Small Business Program: 15% on all sales if you earn under $1M/year. Almost all indie developers qualify; you have to opt in.
- Reader-app exception (specific media categories): allows external purchase flows with reduced commission.
- EU Digital Markets Act alternative distribution (sideloading, third-party stores) is available with separate fee structure — nuanced and changing; check current state if relevant.
The Small Business Program changes the math
Under the Small Business Program, your effective revenue per dollar of gross sale is $0.85 instead of $0.70. On a $9.99 subscription that's $8.49 vs $6.99 — 21% more revenue. For an indie shipping a few apps under $1M total revenue, this is free margin. Apply for it.
Monetization patterns that work on iOS
- Free + freemium unlock (RDR2 Companion's $2.99 unlock pattern). Lowest friction. Best when value is unlocked at a discrete point.
- Free trial → auto-renew subscription. 7-day or 30-day trial, then monthly/annual. Works when value compounds with use (productivity, fitness, learning apps).
- Tiered subscriptions. Pro / Premium tiers for power users. Common in AI apps.
- Consumable IAP (credit packs). Matches AI usage well: users buy credits, credits buy AI calls.
- One-time purchase. $0.99-$4.99 typical. Simple, honest, hard to grow.
- Hybrid. Free with optional cosmetic IAP, plus a "Pro" annual. Many successful apps use this layered approach.
AI cost per iOS user
For a Claude-powered iOS app, AI cost dominates variable expenses. Estimating from the framework:
- Casual user (4 queries/month on Sonnet): ~$0.05/mo AI cost.
- Engaged user (20 queries/month on Sonnet, mostly cached system prompts): ~$0.25-$0.50/mo.
- Power user (100+ queries/month): $1-$3/mo.
- Abusive user (rate-limit them): would be $10-$50/mo without limits.
This is why rate-limiting is non-negotiable and why mixing Haiku (cheap routine answers) with Sonnet/Opus (hard questions) cuts cost meaningfully.
Worked example: RDR2 Companion v1.2 freemium
Assumptions: $2.99 one-time unlock, 4 free AI questions before paywall, target conversion rate 5% of installs to paid.
- Per paid user gross revenue: $2.99.
- Apple Small Business cut (15%): -$0.45 = $2.54 net.
- AI cost for the 4 free queries: ~$0.01 per user (across all installs, including non-converters).
- AI cost for ongoing engagement post-unlock: avg ~$0.15/mo per converted user, ~$1.80 over a year of moderate use.
- Backend/infra amortized per converted user: negligible.
- Net contribution per paid user over a year: ~$0.50 after AI costs (since paid is one-time, AI keeps going).
Tight. Two clear improvements: add a Pro tier ($9.99/year for unlimited Opus-level answers) for power users; consider switching to subscription instead of one-time unlock as users mature into the app.
Hidden costs of iOS
- App review delays. 24-48 hours typically, sometimes longer. Plan launches with buffer.
- Rejection cycles. First-time submissions are often rejected for guideline issues. Budget 1-2 review cycles.
- Refunds. Apple's refund rate runs 0.5%-2% of paid transactions. Real cost.
- Mac for development. You need a Mac. Newer Macs are expected.
- Annual paid developer fee. $99/year.
- Support time. iOS users expect responsive support; budget 1-3 hours/week per ~1,000 paid users for replies.
- Compliance work. Privacy Manifest, App Tracking Transparency, App Privacy disclosures, GDPR/CCPA — ongoing work.
Optimization levers
- Enroll in Small Business Program. Free 15% margin.
- Mix models on the backend. Haiku for cheap queries, Sonnet for harder ones.
- Aggressive prompt caching. Long system prompts cached can cut AI cost 70-90%.
- Rate limits. Hard caps per user/day prevent abuse.
- Annual subscription over monthly. Lower churn, front-loaded cash.
- Pricing experiments. Test 50% higher prices; often you'll find conversion drops less than revenue rises.
- App Store Optimization (ASO). Free traffic from search beats paid ads.
- Push notification re-engagement. Brings users back at zero variable cost.
See: App Economics Framework, iOS App Build Cost, AI Subscription Economics.