For a decade, the UA playbook for mobile games has been written within the defined boundaries of Apple’s App Store and the Google Play Store. The formula was clear: spend on ads, drive to the store, and accept a deterministic 30% commission on all subsequent revenue as the inevitable “platform tax.”
However, a quiet disruption by industry giants has now reached a tipping point. Studios like MiHoYo (Genshin Impact) and Supercell are no longer solely relying on the walled gardens for monetization. They are actively driving users to proprietary Direct-to-Consumer (D2C) web storefronts to process transactions. According to a 2024 report, 26% of players are already open to visiting a publisher’s D2C website, and 14% have made a purchase on one, a trend expected to grow significantly by 2026.
This is not merely a technical adjustment; it is a structural shift in mobile game unit economics. This shift demands a radical overhaul of how UA teams calculate ROAS, model LTV, and approach the entire concept of multi-channel attribution.
In 2024, 16% of UA spend flowed through web-to-app funnels, signaling a shift away from store-only acquisition.
The Rise of the D2C Storefront
The catalyst for this trend is a “perfect storm” of legal pressures, regulatory changes (such as the EU’s Digital Markets Act), and privacy initiatives (like Apple’s App Tracking Transparency). These factors have emboldened developers to challenge the 30% commission, arguing it stifles innovation and anti-competitively reduces margins.
When publishers utilize D2C storefronts, they bypass the platform-controlled payment pipes. The economic incentive is clear: where App Stores take 15%–30%, third-party web payment processors typically charge between 2% and 6% after blending all transaction fees. For a high-grossing title, reclaiming this ~24% margin delta is the difference between a struggling portfolio and a scaling powerhouse.
The D2C approach offers secondary benefits that are equally valuable in the post-ATT era. Web funnels allow developers to collect first-party data before the user ever engages with the app itself. This data restores a level of granularity and ownership that has been lost on mobile native platforms. Furthermore, developers have total “monetization freedom” on the web, allowing them to test bundles, loyalty programs, and pricing tiers that are prohibited by strict App Store guidelines.
How Web Revenue “Inflates” Effective ROAS
The standard definition of Gross ROAS used by UA teams doesn’t fully work in a D2C world. Gross ROAS typically compares ad spend to revenue reported by the App Store. It does not account for the revenue collected through the web shop.
To scale effectively, UA teams must transition to a Net ROAS model that incorporates all payment channels. When web revenue is accounted for, the “effective” ROAS can appear to “inflate” compared to historical app-only benchmarks, even without changing the ad creative or targeting.
A Conceptual Example of the Margin Paradox
Imagine a casual game acquires a cohort of 1,000 users with a CAC of $1,000.
- Historically, these users might spend $1,000 on IAP inside the app.
- The UA manager sees 1.0x Gross ROAS in their dashboard.
- However, the publisher only receives $700 (Net-of-30%-Tax), making the campaign unprofitable.
Now consider the D2C model, where the publisher converts 30% of that cohort’s value to the web shop (through targeted loyalty incentives).
- The users still spend $1,000 total.
- App Store Revenue: $700 (Publisher Net: $490).
- Web Store Revenue: $300 (Publisher Net at 6% fee: $282).
- Total Net Publisher Revenue: $772.
The Gross ROAS still looks like 1.0x (spend to gross spend), but the Effective Net ROAS has increased by ~10%. The campaign is now close to breaking even on cash flow, solely due to the payment “pipe” chosen by the user. UA teams that fail to account for this delta are undervaluing their cohorts and prematurely killing profitable campaigns.
The New LTV Model: Why Web Revenue is Non-Negotiable
The paradigm shift from Gross to Net economics dictates that UA teams must incorporate web-based revenue streams into their LTV forecasting models.
Traditionally, LTV is modeled as:
LTV = (ARPDAU * Lifetime) / Total Users.
In a multi-channel world, this formula is incomplete. A modern LTV model must include a Multi-Channel Multiplier based on the probability of a user migrating to the web shop. This multiplier is not uniform across all users; it segments based on cohort propensity.
The “Web-Propensity Multiplier”
Top UA teams are now segmenting cohorts not just by geo or OS, but by their Web-Propensity Score. If a campaign on TikTok is successfully acquiring “whales” who are highly incentivized to use the discounted D2C web shop for large currency purchases, that cohort’s LTV is drastically higher than a cohort of “casual spenders” who only make small, friction-free $0.99 purchases inside the native app.
By factoring the higher margins of web revenue into their LTV models, UA managers can justify a higher initial Cost Per Install (CPI) or Cost Per Action (CPA) for high-propensity cohorts, allowing them to outbid competitors who are still modeling LTV based on App Store nets alone.
The Attribution Black Hole: Cross-Platform Challenges and MMP Limitations
While the economics of D2C storefronts are attractive, the technical implementation presents significant hurdles. The primary challenge is the attribution “black hole” created when a user journey breaks between the web ecosystem and the mobile app ecosystem.
Cross-Platform Tracking
Tracking a user from a “web-to-app” ad flow (e.g., clicking a web ad that leads to a web onboarding quiz, then to a web purchase, and finally to an app install) is complex. Identify consistency is the major barrier. Walled gardens (Meta, Google, TikTok) use their own proprietary identifiers, making it difficult to “stitch” together a user who clicks a Meta ad and later searches on Google.
Mobile Measurement Partner (MMP) Limitations
Most MMPs were designed to track native mobile app events. They provide a unified view of mobile-specific metrics (installs, reattributions, SKAN conversions), but often struggle to ingest and merge web-based purchase data in real-time.
When a user jumps from a browser to a native app, the connection is frequently broken. While top MMPs like Adjust and AppsFlyer offer solutions for web attribution and deep-linking, preserving continuity across these flows requires flawless deep-linking implementation. If the connection is lost, a UA dashboard might report 0% ROAS for a campaign that actually drove tens of thousands of dollars in high-margin web revenue.
Final Thoughts: Adapting Campaigns and Attribution to Web-to-App Flows
The era of monolithic App Store dominance is kind of over for major mobile publishers. The 30% commission is no longer a necessary evil, but an adjustable variable.
UA teams must adapt by treating web-to-app funnels as a core pillar of their strategy, rather than an experimental edge case. To thrive in this new landscape, teams must:
- Transition to “Net ROAS” KPIs: Stop measuring campaigns against gross revenue. Create unified dashboards that automatically net out both 30% (App Store) and ~6% (Web) transaction fees to show true cash-flow profitability.
- Rethink LTV Models: Incorporate segmented multipliers for web conversion probability. Adjust CPI bids based on a cohort’s modeled propensity to bypass the native store.
- Bridge the Attribution Gap: Demand more robust multi-channel integration from MMPs. Implement server-to-server (S2S) purchase postbacks to unify all revenue data, allowing web transactions to be correctly attributed back to the original acquisition source (the ad creative).
The studios that master the math of multi-channel monetization will find themselves with a powerful competitive advantage: the ability to acquire high-value users at a lower net cost, while maximizing the total value extracted from every install.
