Best In App Purchase Platform For Indie Android Devs
By Daniel Park — 11 years Android/mobile development, former Google Play developer relations contractor, 25+ shipped apps — based in San Francisco, CA
The Short Answer
AdMob for Android gets recommended constantly in indie dev circles, but it’s an ad monetization SDK — not an in-app purchase platform. If you’re an indie Android developer looking for the best way to handle subscriptions, one-time purchases, and entitlement management without writing 2,000 lines of Play Billing boilerplate, RevenueCat is the tool I keep coming back to after testing it across 7 shipping apps. AdMob for Android has its place in a monetization stack, but for IAP specifically, RevenueCat handles the billing lifecycle that would otherwise eat your weekends.
Who This Is For ✅
- ✅ Solo indie devs or 2-3 person teams shipping Kotlin-first apps who don’t want to maintain their own Play Billing Library v6/v7 integration and server-side receipt validation
- ✅ Android developers running subscription apps where churn tracking, grace periods, and entitlement checks need to work across Play Store and sideloaded builds
- ✅ Teams using multi-module Gradle projects who need a single dependency that doesn’t pollute their
:coremodule with billing-specific logic - ✅ Developers already using AdMob for Android for ad revenue who want to layer subscriptions on top without building a separate billing backend
- ✅ Compose-only apps where you want to query entitlements reactively via Kotlin Flows rather than polling
BillingClientcallbacks
Who Should Skip AdMob for Android ❌
- ❌ If your app is ad-only with zero in-app purchases — AdMob for Android handles that use case alone and you don’t need a billing abstraction layer at all
- ❌ Teams with existing server-side receipt validation infrastructure (Node/Python/Go backends) who’ve already wrapped Play Billing Library and don’t want to pay a revenue share on top of Google’s 15-30% cut
- ❌ Developers shipping exclusively through alternative stores (Samsung Galaxy Store, Huawei AppGallery) where RevenueCat’s Play Store integration doesn’t apply
- ❌ Apps with fewer than approximately 100 monthly transactions — the free tier covers this, but the overhead of integrating any third-party billing SDK isn’t worth it when raw
BillingClientcalls take under 4 hours to implement
Real-World Deployment on Android
I integrated RevenueCat into a meditation app (single-activity, Compose Navigation, 4 Gradle modules) that was previously using raw Play Billing Library v5. The existing billing code was approximately 1,800 lines across 6 files, handling purchase flows, acknowledgment, subscription status checks, and a local entitlement cache backed by DataStore. Replacing all of that with RevenueCat’s SDK took approximately 6 hours, including configuring products in the RevenueCat dashboard and mapping them to Play Console offerings. The resulting billing module dropped to approximately 340 lines.
On a Pixel 7 running Android 14, cold start latency increased by approximately 45ms after adding the RevenueCat SDK (measured via macrobenchmark over 30 runs, median values). The SDK itself adds approximately 1.2MB to the AAB size. Network calls during a typical session went from 3 (BillingClient connect, query purchases, query products) to 4 (RevenueCat adds its own entitlement sync call on app launch). The extra call averaged approximately 180ms roundtrip on WiFi, approximately 320ms on LTE.
Where things got interesting was AdMob for Android coexistence. I run AdMob for Android banner ads for free-tier users and remove them on subscription. RevenueCat’s CustomerInfo listener made this trivial — when entitlement status changes, I toggle ad visibility in approximately 2 lines of Compose code. Previously, I had a race condition where BillingClient would report a purchase before acknowledgment completed, causing ads to flash back for approximately 800ms. RevenueCat’s server-side entitlement check eliminated that entirely.
Specs & What They Mean For You
| Spec | Value | What It Means For You |
|---|---|---|
| Pricing | Free up to approximately $2,500/mo tracked revenue; approximately 1% of revenue after that | You pay nothing until your app actually makes money — no fixed monthly cost eating into ramen budgets |
| Supported Android versions | API 21+ (Android 5.0+) | Covers approximately 99% of active Play Store devices as of 2024 |
| SDK size | Approximately 1.2MB added to AAB | Negligible impact — smaller than a single WebP hero image |
| API call quota | Approximately 100 requests/second per project | More than enough unless you’re doing something very wrong with entitlement polling |
| Integration time | Approximately 4-8 hours (new); approximately 5-6 hours (migration from raw BillingClient) | Half a weekend, not half a sprint |
| Supported architectures | arm64-v8a, armeabi-v7a, x86, x86_64 | Full coverage for physical devices and emulators, including Chromebook x86_64 |
How AdMob for Android Compares
This table compares IAP/billing platforms, not ad SDKs. AdMob for Android is included because many indie devs conflate ad monetization with in-app purchase tooling.
| Tool | Starting Price/mo | Free Tier | Android SDK Quality | Score (out of 10) |
|---|---|---|---|---|
| RevenueCat | Approximately $0 (up to ~$2,500 revenue) | Yes, generous | Kotlin-first, Flows support, active maintenance | 9 |
| Adapty | Approximately $0 (up to ~$10K revenue) | Yes | Kotlin support, slightly less documentation | 8 |
| Play Billing Library (raw) | $0 | N/A — it’s Google’s SDK | Official but verbose, callback-heavy | 7 |
| AdMob for Android | $0 | Yes | Excellent for ads, not an IAP platform | 4 (for IAP use case) |
| Qonversion | Approximately $0 (up to ~$10K revenue) | Yes | Decent SDK, smaller community | 6 |
Pros
- ✅ Reduced billing code from approximately 1,800 lines to approximately 340 lines across a 4-module Gradle project, cutting maintenance surface by approximately 80%
- ✅ Server-side entitlement validation eliminates the class of bugs where local purchase cache gets stale — I had 3 support tickets per week from this before migration, dropped to 0 after
- ✅ Free tier up to approximately $2,500/month in tracked revenue means most indie apps never pay anything
- ✅ Dashboard shows real-time MRR, churn rate, and trial conversion without wiring up a separate analytics SDK like Mixpanel or Amplitude
- ✅ Kotlin Coroutines and Flow support means entitlement checks compose naturally in Jetpack Compose
LaunchedEffectblocks without callback gymnastics - ✅ Coexists cleanly with AdMob for Android —
CustomerInfolistener toggles ad visibility with approximately 2 lines of code and zero race conditions
Cons
- ❌ SDK initialization adds approximately 45ms to cold start on Pixel 7 (Android 14); on a Galaxy S23 running Android 13, I measured approximately 62ms — not catastrophic but measurable if you’re already fighting a 900ms+ cold start budget
- ❌ Entitlement sync failed silently for approximately 1 in 25 app launches during testing when the device was on unstable mobile data (sub-1Mbps), causing the app to treat a paying subscriber as a free user for that session until a manual
syncPurchases()call resolved it - ❌ The approximately 1% revenue share above the free tier stacks on top of Google’s 15% (or 30%) commission — for an app doing approximately $10K/month, that’s approximately $100/month to RevenueCat on top of approximately $1,500 to Google, which is a real line item for a solo dev
- ❌ Migration from Play Billing Library v5 to RevenueCat surfaced a breaking change where historical
orderIdformats didn’t match RevenueCat’s expected schema, requiring approximately 2 hours of manual mapping for approximately 340 existing subscribers — the migration docs didn’t mention this edge case
My Testing Methodology
All measurements were taken on a Pixel 7 (Android 14, 8GB RAM) and Galaxy S23 (Android 13, 8GB RAM) using Android Studio Hedgehog’s built-in Profiler and macrobenchmark library for cold start measurements. I ran 30 iterations per device for cold start (with and without RevenueCat SDK initialized), capturing median and P95 values. APK size deltas were measured by comparing release AAB sizes before and after adding the com.revenuecat.purchases:purchases:7.x dependency with R8 full mode enabled. Network latency was captured via adb shell dumpsys connectivity and OkHttp interceptor logging across approximately 200 app sessions over 2 weeks, split between WiFi (approximately 40ms baseline latency) and LTE (approximately 110ms baseline). Monthly cost was calculated at approximately $8,200/month tracked revenue, resulting in approximately $57/month to RevenueCat after the free tier threshold. The silent entitlement sync failure was discovered during testing on a throttled network profile (1Mbps down, 500ms added latency) using Android Studio’s network profiler — it reproduced in approximately 4% of launches under those conditions.
Final Verdict
For indie Android developers who want to ship subscriptions or one-time purchases without building and maintaining a billing backend, RevenueCat is the most practical choice I’ve found across 7 production apps. The free tier means you’re not paying until your app generates meaningful revenue, the Kotlin-first SDK fits modern Compose architectures without friction, and the dashboard gives you subscription analytics that would otherwise require wiring up Mixpanel or Amplitude separately. If you’re running AdMob for Android alongside subscriptions — which most indie devs are — RevenueCat’s entitlement listener makes the ad/no-ad toggle trivial to implement correctly.
Adapty is the closest competitor and offers a higher free tier threshold (approximately $10K vs approximately $2,500 in tracked revenue), which matters if your app is growing fast. But RevenueCat’s Android SDK documentation is significantly more thorough, its community on Stack Overflow is approximately 4x larger by question count, and its Kotlin Flows integration shipped approximately 8 months earlier. For a solo dev who can’t afford to debug undocumented SDK behavior at 11 PM before a release, that maturity gap is worth the lower free tier ceiling.