--- title: Measuring the ROI of ASO and featuring slug: aso-roi-measurement summary: How in-house teams (Nordeus, Tilting Point, Zynga) measure ASO — installs as the north star, holistic vs single-activity ROI, a D7→D365 LTV multiplier when there's no predictive model, the featuring-uplift calculation, and using App Store Connect cohorts and unified tracking links. domain: mobile-apps tags: [aso, measurement, roi, ltv, featuring, app-store-connect, attribution] related: [aso-audit-and-prioritization, paid-to-organic-uplift, in-app-events-and-promotional-content, aso-as-acquisition-channel, measurement-stack-attribution-incrementality-mmm, incrementality-test-design, measuring-organic-creator-attribution] sources: [yt:nE6i24tLAeI, yt:3nas91cJjDE, yt:JtvOFVzxVLM, yt:M2e6pJuA8hA] updated: 2026-10-24 --- # Measuring the ROI of ASO and featuring ## TL;DR - **Installs are the north star** for organic; conversion rate, impressions, page views and organic revenue are helper metrics.[^1][^2] - Conversion rate alone **misleads** — product teams ask for it, but Zynga's ASO lead steers them to install growth.[^3] - Two ROI views: **holistic** (all marketing combined, split organic vs paid) or **focused** (one activity, where correct cost allocation matters most).[^4][^5] - No LTV model? Project 1-year value as **D7 LTV × your historical D7→D365 multiplier** — Nordeus found it stable over time.[^6][^7] - **Featuring uplift** = (daily average during featuring − clean baseline − installs from other concurrent activity) × days featured, then × projected LTV.[^8][^9] - Don't trust console "organic" blindly — on Google Play part of it is paid (Google Ads/UAC) traffic.[^10] ## Details **Holistic ROI (Aleksandra Zivkovic, Nordeus/Top Eleven, 2022).** Split new registrations into organic and paid as best your internal analytics allow (never perfect), multiply each by expected revenue per user from a predictive LTV model, allocate costs to each, and compute ROI for the period.[^4][^11][^12] **Without predictive modelling.** Take actual D7 LTV of the cohort and apply the D7→D365 multiplier observed on historical cohorts; Aleksandra found these multipliers move little unless something extraordinary happens.[^6][^7] If you build an LTV model, make it useful for other marketing disciplines too, not just ASO.[^13] **Worked example: was App Store featuring worth pitching for?**[^8][^9] 1. Daily average organic registrations over the 4 featured days. 2. Baseline: comparable days with few other activities, chosen with seasonality in mind. 3. Subtract installs attributable to concurrent activities (in her example, an influencer campaign = 100 installs). 4. Daily uplift × featured days = total registration uplift. 5. × (featuring cohort D7 LTV × historical multiplier) = estimated 1-year revenue uplift. Storemaven's Jonathan Fishman (in conversation with AppTweak's Simon Thillay, 2022) suggests the same from App Store Connect: cohorted in-app purchase data lets you isolate the day-of-featuring cohort and put a dollar value on it, and CPP-level revenue gives aggregated campaign ROAS ([[custom-product-pages-for-ads]]).[^14][^15] **Tooling.** Tilting Point (30+ games, 2022) pulls both consoles via API into a Looker dashboard so a person covering several games can scan them in minutes, not a day.[^16] Simon notes keyword search volume doubles as a brand-awareness gauge: compare your brand term's volume with competitors'.[^17] **Falling conversion is often good news (Shoham Lashkar, Zynga, 2023).** A brand searcher may convert at ~100%; as you rank for more generic terms, page traffic grows with lower intent and CR drops. He asks product teams whether they prefer 100% of 100 brand searchers or 20% of a million — compare CR only inside a test with bought traffic, and otherwise check that installs rise with reach.[^21][^22] Apple Ads data fills a gap the consoles leave: it shows conversion and revenue per keyword, which tells you which organic ranks are worth chasing.[^23] **Unify tracking links (Amy Monier, Zynga, 2023).** Put the same MMP (AppsFlyer/Adjust) deep links on social pages, email and in-product referral invites; otherwise social and player-to-player installs are lumped into "organic/unknown" and those teams get no credit. Referral tracking also shows invite value, so you can reward inviters accordingly.[^18][^19][^20] ## Disagreements / caveats - Every method here is uplift/correlation, not a controlled experiment; for budget decisions see [[incrementality-test-design]] and [[measurement-stack-attribution-incrementality-mmm]]. - Steve P. Young ignores App Store Connect "impressions" entirely and watches page-view→download conversion ([[aso-ranking-factors]]); Shoham Lashkar instead downplays conversion rate in favour of total installs — both agree impression-based CR is a weak headline metric.[^3] ## See also - [[paid-to-organic-uplift]] — estimating how many organic installs paid spend creates - [[aso-audit-and-prioritization]] [^1]: yt:3nas91cJjDE @ 04:55 — Jevgen Tarasenko [^2]: yt:JtvOFVzxVLM @ 29:59 — Shoham Lashkar [^3]: yt:JtvOFVzxVLM @ 30:56 — Shoham Lashkar [^4]: yt:nE6i24tLAeI @ 07:41 — Aleksandra Zivkovic [^5]: yt:nE6i24tLAeI @ 12:54 — Aleksandra Zivkovic [^6]: yt:nE6i24tLAeI @ 11:17 — Aleksandra Zivkovic [^7]: yt:nE6i24tLAeI @ 12:08 — Aleksandra Zivkovic [^8]: yt:nE6i24tLAeI @ 14:32 — Aleksandra Zivkovic [^9]: yt:nE6i24tLAeI @ 15:26 — Aleksandra Zivkovic [^10]: yt:3nas91cJjDE @ 05:42 — Jevgen Tarasenko [^11]: yt:nE6i24tLAeI @ 09:28 — Aleksandra Zivkovic [^12]: yt:nE6i24tLAeI @ 10:23 — Aleksandra Zivkovic [^13]: yt:nE6i24tLAeI @ 16:23 — Aleksandra Zivkovic [^14]: yt:M2e6pJuA8hA @ 19:51 — Jonathan Fishman [^15]: yt:M2e6pJuA8hA @ 15:29 — Jonathan Fishman [^16]: yt:3nas91cJjDE @ 28:38 — Jevgen Tarasenko [^17]: yt:M2e6pJuA8hA @ 22:59 — Simon Thillay [^18]: yt:JtvOFVzxVLM @ 53:28 — Amy Monier [^19]: yt:JtvOFVzxVLM @ 54:44 — Amy Monier [^20]: yt:JtvOFVzxVLM @ 55:59 — Amy Monier [^21]: yt:JtvOFVzxVLM @ 31:30 — Shoham Lashkar [^22]: yt:JtvOFVzxVLM @ 32:34 — Shoham Lashkar [^23]: yt:JtvOFVzxVLM @ 33:59 — Shoham Lashkar