Deep Dive on Two Key Uncertainties: App Store Fees & Implementation Timelines — March 14, 2026
Methodology: David Deutsch's criterion — only non-variable explanations that are hard to change without also changing the observed facts. 2–3 core explanations per section maximum.
| Scenario | Likely Fee Range | Rationale | Probability |
|---|---|---|---|
| Bull case | 0–5% | Court interprets "costs" narrowly — Apple's actual infrastructure cost for enabling a link is near-zero. Epic/Sweeney celebrated ruling as "death of Apple Tax." | ~30% |
| Base case | 5–12% | Court grants some IP value for iOS platform + App Store distribution. Aligns with EU DMA structure (5% CTC + some store services). Google's User Choice Billing charged ~4% less than standard. | ~50% |
| Bear case | 12–20% | Court agrees Apple's IP value is substantial. Still materially below 30% but not the windfall developers hoped for. | ~20% |
Current state: Zero commission on external links (contempt order still in effect while "reasonable fee" is being determined).
Future state: Some reasonable fee TBD by district court. Could take 6–12 months to resolve.
Can Apple block web D2C? No. The injunction is permanent and upheld by the 9th Circuit.
Current state: Under DMA, apps can use external payment and link out. Fee structure: ~10–20% total depending on path (5% CTC + Store Services + Acquisition fee).
Future state: Jan 2026 single business model transition. European Commission actively monitoring compliance.
Can Apple block web D2C? No. DMA Article 5(4) prohibits it. EC has enforcement powers with fines up to 10% of global revenue.
| Company | When | What They Did | Timeline | Outcome |
|---|---|---|---|---|
| Netflix | Dec 2018 | Stopped accepting new iOS IAP signups. Directed all new users to web. Already had web billing infrastructure from day one (Netflix was web-first). | ~0 months (already had web billing). Took until Feb 2024 to fully cut off grandfathered IAP subscribers. | No meaningful churn reported. Users already had Netflix accounts with web credentials. |
| Spotify | May 2016 | Stopped accepting new iOS IAP signups after only 2 years of offering it (Jun 2014–May 2016). Directed to web. | ~0 months (Spotify was web-first). Cut off remaining IAP subscribers in Jul 2023. | By May 2025, immediately added in-app external link post-ruling — took ~1 day to submit update. |
| Epic/Fortnite | Aug 2020 | Added direct payment option in-app (bypassing IAP). Already had web store and Epic Games Store ecommerce. | Built direct payment in ~months (had ecommerce infra). Launched "Web Shops" for third-party devs Jun 2025. | Got removed from App Store (triggered lawsuit). V-Bucks web purchases worked fine technically. |
| Match Group / Tinder | 2019–2022 | Added direct credit card entry in Android app (2019). Fought Google on alternative billing. Put $40M in escrow. | 6–12 months for Android direct billing. Multi-year legal battle for compliance. | Proven that dating apps can do D2C. Match has been doing it on Android for years. |
| Bumble | Q3 2025–present | "As soon as permitted" after Apr 2025 ruling, started testing alternative payments. Rolled out Apple Pay direct billing in Q4 2025. | ~3–4 months from ruling to testing (Q3). ~6 months to measurable margin impact (Q4). | 1pp gross margin expansion in Q4 2025. Apple Pay adoption >50% of US iOS payments by early Q1 2026. |
This question is partially obsolete because Bumble is already doing it and reporting results:
| Data Point | Source |
|---|---|
| Direct billing testing began Q3 2025 — "as soon as permitted" | Bumble Q3 2025 earnings call (Nov 6, 2025) |
| Nearly all US members had some form of direct billing by Q3 | Alpha-Sense Q3 2025 summary |
| Apple Pay program implemented in Q4 2025 | Bumble Q4 2025 earnings call (Mar 11, 2026) |
| 1 full percentage point of Y/Y gross margin expansion from alternative billing in Q4 | CFO Kevin Cook, Q4 2025 call |
| Apple Pay adoption >50% of US iOS payments by early Q1 2026 | Bumble Q4 2025 earnings call |
| "Benefits expected to increase through 2026" | Multiple Q4 2025 transcript sources |
Redirecting users from seamless 1-tap IAP to web checkout = friction = lower conversion.
IAP has Face ID / Touch ID / saved payment. Web requires login + card entry.
Bumble used Apple Pay as the D2C method — not a redirect to web. Same biometric tap, lower commission.
>50% adoption rate in months suggests minimal friction. Users don't even notice the difference.
This is the key insight: D2C ≠ clunky web checkout. Apple Pay / Google Pay direct billing preserves UX.
| Solution | What It Does | Fee | Timeline to Integrate |
|---|---|---|---|
| Stripe Billing | Full subscription management, recurring billing, dunning, analytics. No-code migration toolkit for importing existing subscriptions. | 2.9% + $0.30 per transaction | Weeks for basic; 2–3 months for full migration with entitlement sync |
| RevenueCat Web Billing | Cross-platform subscription management. Web SDK that syncs entitlements with iOS/Android apps. "No coding or hosting required." | $0.012/MTR (monthly tracked revenue per $1) | Days for basic web paywall; weeks for full cross-platform sync |
| Adapty | Paywall optimization, A/B testing, web+mobile subscription unification. | Tiered pricing | Weeks |
| Epic Web Shops | White-label web storefronts for game developers. Launched Jun 2025. | 12% (Epic takes) | Self-service setup |
The ecosystem of D2C billing infrastructure has matured dramatically since 2025. What once required custom engineering is now largely turnkey.
The D2C shift creates a structural reallocation of ~$20B+ in annual app store commissions toward payment processors, billing platforms, and subscription management tools. The question is who captures that flow.
| Company | Ticker | Why They Win | Exposure |
|---|---|---|---|
| Stripe | Private | Default payment processor for web. Stripe Billing has purpose-built subscription migration tools. Every app that leaves IAP for web likely lands on Stripe. | 🔴 High — this is their core business |
| Adyen | ADYEN (AMS) | Enterprise payment processing. Already handles payments for Spotify, Netflix, Uber. Natural recipient of large D2C migrations. | 🟠 Medium-High — large enterprise focus |
| Block (Square) | XYZ | Payment infrastructure, but more SMB/physical. Less direct D2C app exposure than Stripe/Adyen. | 🟡 Medium |
| Company | Status | Why They Win |
|---|---|---|
| RevenueCat | Private (Series C, $1.5B+ valuation) | The "picks and shovels" play. Already the dominant mobile subscription SDK. Now has web billing. Every developer exploring D2C touches RevenueCat. Their web billing product launched specifically for the post-Epic world. |
| Chargebee | Private | Enterprise subscription billing. Handles complex pricing, dunning, revenue recognition. Target for larger companies building D2C. |
| Adapty | Private | Mobile paywall optimization + web. Competes with RevenueCat on the experimentation/optimization layer. |
| Category | Examples | Mechanism |
|---|---|---|
| Identity / Auth | Okta, Auth0 | Web checkout requires user authentication outside app stores. More web = more auth flows. |
| Fraud Prevention | Sift, Riskified | App stores handled fraud. D2C shifts fraud liability to developers, who need third-party solutions. |
| Analytics / Attribution | Amplitude, Mixpanel | Tracking cross-platform purchase funnels is harder without IAP. More demand for analytics tooling. |
| Question | v2 Assessment | v3 Assessment (Now) |
|---|---|---|
| Can Bumble do D2C? | Likely yes, post-Epic ruling | Already doing it. 1pp margin impact. >50% Apple Pay adoption. |
| Will Apple block it? | Unlikely given court rulings | Cannot. Injunction permanent. Only question is fee level (0–12%). |
| How long to implement? | Uncertain — 6–18 months? | ~3–6 months to measurable impact. Already done. |
| Conversion rate risk? | Moderate concern | Minimal. Apple Pay preserves UX. >50% adoption proves it. |
| Reasonable fee risk? | Not analyzed in v2 | NEW RISK: Appeals court says Apple can charge something. Base case 5–12%. Even worst case, still saves 10pp+ vs IAP. |
Sources: Bumble Q3 2025 & Q4 2025 earnings call transcripts (Motley Fool, Investing.com, Globe & Mail); MacRumors (Dec 11, 2025 appeals ruling); Ars Technica; TechCrunch (May 2, 2025 Apple guidelines update); RevenueCat blog; Apple Developer documentation; The Verge; Variety; Reuters.
Report date: March 14, 2026 | Methodology: Non-variable explanations per David Deutsch