Ios app growth marketing

iOS App Growth & Marketing

End-to-end growth engineering for iOS/iPadOS apps. This skill covers the four pillars of app growth and the iOS-native plumbing each one depends on:

1. Acquisition — paid UA, ad-attribution measurement under ATT (the iOS-specific part most generic marketing advice gets wrong).
2. Monetization — subscription/IAP pricing strategy, StoreKit 2 + RevenueCat paywall pricing display, purchasing-power-parity territory pricing.
3. Activation & retention — onboarding-to-paywall flow, CRO, referral/viral loops, churn reduction.
4. Strategy — launch sequencing, growth-tactic selection, marketing psychology, competitor positioning.

Scope boundary. This skill is about growing an app that already exists. It does not cover App Store listing optimization (keywords, screenshots, title/subtitle — that's the app-store-optimization / ASO skill), the release/submission mechanics (asc CLI, TestFlight, build upload — that's app-store-release), or App Review compliance checks (that's app-review-preflight). When a request is purely ASO, release, or review, delegate. This skill owns demand generation, attribution measurement, pricing, CRO, and viral/retention loops.

---

Decision tree — route to the right reference

What does the user need?
│
├─ "Installs going Organic / attribution missing / ATT prompt / SKAN / AdAttributionKit
│   / conversion value / Meta AEM / MMP setup / IDFA / postback receiver / dual-call"
│   └─→ references/attribution-and-att.md
│       The iOS-native acquisition-measurement layer. SDK init order, ATT 4-state
│       handling, SKAdNetwork v4 + AdAttributionKit (incl. iOS 18 PostbackUpdate +
│       server-side postback receiver), per-channel measurement, MMP troubleshooting.
│
├─ "TikTok SDK / TikTokBusinessSDK / disableSKAdNetworkSupport / tiktokAppId / SKAN owner"
│   └─→ references/tiktok-channel.md
│       Business-SDK-vs-OpenSDK trap, SKAN-ownership decision matrix, init order,
│       hybrid event_id dedup, Events API S2S.
│
├─ "Meta AEM greyed-out / extinfo / ATE verification error / CAPI app events / CUID"
│   └─→ references/meta-attribution.md
│       7-prereq AEM checklist, ATE 5-cause list, the extinfo 16-position array.
│
├─ "Apple Search Ads / ASA / CPT / keyword bids / campaign structure / Search Terms / CPP"
│   └─→ references/apple-search-ads.md
│       Readiness gate, CPT-by-category, bid formula, CPP routing, 2025-26 changes.
│
├─ "In-App Events / Today tab event / event card / live event / challenge"
│   └─→ references/in-app-events.md   (free discovery + no-opt-in re-engagement push)
│
├─ "App Clip / instant experience / SKOverlay / App Clip card / 3-5x conversion"
│   └─→ references/app-clips-growth.md
│
├─ "Benchmarks / retention bands / D1/D7/D30 / cohort analysis / LTV by market / churn"
│   └─→ references/metrics-benchmarks.md
│       Per-category Poor/Avg/Good/Excellent bands, cohorts, sourced 2026 sub data.
│
├─ "Rating prompt / requestReview / star rating / pre-prompt survey / rating reset"
│   └─→ references/ratings-and-reviews.md
│
├─ "Paid ads / ROAS / CPA / campaign structure / ad creative / audiences / retargeting"
│   └─→ references/paid-acquisition.md
│       Platform selection, campaign structure, ad copy/creative frameworks,
│       audience targeting, optimization levers, blended-CAC accounting.
│
├─ "Pricing / subscription tiers / freemium / free trial / willingness to pay /
│   value metric / price increase / Van Westendorp / MaxDiff / unit economics"
│   └─→ references/monetization-and-pricing.md
│       Value-based pricing, research methods, tier structure, freemium vs trial,
│       app subscription unit economics, when/how to raise prices.
│
├─ "Paywall price display / RevenueCat / StoreKit price / currency / trial copy /
│   savings badge / hardcoded $ / per-month equivalent"
│   └─→ references/paywall-pricing-display.md
│       The hard rule: prices come from StoreKit/RevenueCat, never Swift literals.
│       Locale-correct price formatting, derived savings %, trial labels.
│
├─ "PPP / purchasing power parity / per-territory pricing / regional pricing /
│   price import / storefront equalization"
│   └─→ references/ppp-territory-pricing.md
│       `asc` CLI flows for subscription + IAP territory pricing via CSV import.
│
├─ "Onboarding / questionnaire flow / activation / CRO / paywall conversion /
│   permission priming / signup flow"
│   └─→ references/onboarding-and-cro.md
│       The 14-screen high-converting onboarding framework, permission priming,
│       paywall placement, CRO levers.
│
├─ "Launch / Product Hunt / go-to-market / waitlist / early access / feature announcement"
│   └─→ references/launch-strategy.md
│       ORB channel framework, 5-phase launch, Product Hunt playbook, ongoing launches.
│
├─ "Referral / affiliate / viral loop / refer-a-friend / K-factor / word of mouth"
│   └─→ references/referral-and-viral.md
│       Referral loop design, incentive sizing, viral-coefficient modeling, fraud.
│
├─ "Marketing psychology / mental models / cognitive bias / persuasion / why people buy"
│   └─→ references/marketing-psychology.md
│       70+ mental models mapped to app marketing applications.
│
├─ "Marketing ideas / growth ideas / how to market / tactics / ways to promote"
│   └─→ references/growth-tactics-library.md
│       140 growth tactics organized by category, stage, budget, timeline.
│
└─ "Competitor pages / vs pages / alternatives / comparison landing pages"
    └─→ references/competitor-pages.md
        Four comparison-page formats, centralized competitor data, SEO + schema.

Read the relevant reference(s) before answering. Each is self-contained.

---

The iOS growth equation (read this first)

App growth compounds three multipliers. Optimize the constraint, not your favorite lever (Theory of Constraints — see references/marketing-psychology.md):

Growth = Installs × Activation × (Monetization − Churn) × Virality

         └ acquisition ┘ └ onboarding ┘ └ pricing ┘    └ referral ┘
         attribution      CRO            StoreKit/RC     K-factor
         tells you which
         installs actually
         work

Why attribution comes first. On iOS post-14.5, if your install measurement is broken, every downstream optimization is flying blind — you can't tell which campaign, creative, or channel produced a paying user, so you scale the wrong spend. The #1 iOS-specific growth bug is "installs going Organic": a marketing SDK fires its install event before the user answers the ATT prompt, so the install is sealed with a zero IDFA and can never be re-attributed. Fix measurement (references/attribution-and-att.md) before scaling acquisition.

---

Core iOS attribution model (post iOS 14.5, current through iOS 26)

Three layers run simultaneously. A healthy stack wires all three; if only one is connected you lose 30-70% of attribution coverage.

┌─────────────────────────────────────────────────────────────────┐
│ Layer 1 — Deterministic (IDFA)                                   │
│   Requires ATT .authorized (~25-50% opt-in, varies by category)  │
│   → full user-level events to AppsFlyer / Adjust / Branch        │
├─────────────────────────────────────────────────────────────────┤
│ Layer 2 — SKAdNetwork v4 + AdAttributionKit (iOS 17.4+)          │
│   Privacy-preserving, no IDFA, Apple-mediated postbacks 24-48h   │
│   Conversion values: fine 0-63 OR coarse low/medium/high         │
│   Apple evaluates SKAN + AAK together; one impression wins       │
├─────────────────────────────────────────────────────────────────┤
│ Layer 3 — Probabilistic / modeled (MMP-side, where Apple allows) │
│   Statistical modeling for non-consented users                   │
└─────────────────────────────────────────────────────────────────┘

The two production failures that account for most iOS attribution loss:

1. ATT wait-scope trap — the MMP's wait API (waitForATTUserAuthorization, attConsentWaitingInterval, enableFirstSessionDelay()) only queues that SDK's events. Backend/CAPI sends, direct Meta/TikTok SDK calls, and events fired before the wait is configured still leak with zero IDFA.
2. Silent consent-denied paths — marketing SDKs early-return without logging when ATT is denied; the dashboard quietly drops 50% of events.

Full diagnosis and code: references/attribution-and-att.md.

---

Monetization quick rules (apps live and die on subscription economics)

• Never hardcode a price in Swift. Prices come from StoreKit (via RevenueCat). A hardcoded $2.50/mo label that drifts from the real charge is an App Review 3.1.2 violation and a refund storm. Full rule: references/paywall-pricing-display.md.
• Price the value, not the cost. Set price between the next-best alternative and the customer's perceived value; cost is a floor. Use Van Westendorp / MaxDiff to find the range. See references/monetization-and-pricing.md.
• Model before you ship — pricing is a one-way door. Once a tier is live and purchased, you can't materially raise it without grandfathering chaos. Verify P95/P99 cost-of-service stays below price after Apple's 15%/30% cut.
• Use PPP for international. Lower-income storefronts get stronger discounts via asc subscriptions pricing prices import (CSV, dry-run first). See references/ppp-territory-pricing.md.
• Onboarding sells the paywall. A 14-screen questionnaire onboarding that surfaces the user's pain, mirrors it back, primes permissions, and lets them use the core mechanic before the paywall converts dramatically better than a cold paywall. See references/onboarding-and-cro.md.

---

Operating sequence for any growth task

1. Identify the pillar (acquisition / monetization / activation / strategy) and route via the decision tree.
2. Gather context the reference lists (product, stage, budget, current metrics — ARPU, CAC, conversion, churn, opt-in rate). Ask if not provided.
3. For audits (e.g. "is our attribution healthy?", "is our pricing right?"): produce a scored PASS / WARNING / FAIL per check, citing the exact code call site (file:line) in the user's project when fixing.
4. For new setups: walk the reference's checklist step by step, gating on each prerequisite.
5. Always reason about second-order effects (a flash sale trains discount waiting; a loose referral reward attracts fraud) — references/marketing-psychology.md catalogs the traps.

---

Per-channel acquisition quick reference (iOS)

Channel	iOS measurement mechanism	ATT dependency	Key SDK
Apple Search Ads (ASA)	AdServices token + AAK/SKAN	Not IDFA-dependent	AdServices API / MMP
Apple Search Ads — keyword strategy	— (delegate to ASO/ASA skill)	—	—
Google App Campaigns	SKAdNetwork + Firebase events	Required for user-level	Firebase + Google Ads
Meta App Install	AEM + SKAN + SDK/CAPI app events	Authorized → IDFA-level signal	Meta SDK / CAPI / MMP
TikTok App Install	SKAN + SDK/S2S events	Required for user-level	TikTok SDK / MMP
Snap App Install	SKAN	Required	Snap SDK / MMP
Branch/Adjust/AppsFlyer/Singular (MMP)	Aggregator across all above + IDFA	Required for user-level	Their SDK

---

What this skill deliberately does NOT do

• App Store listing / keyword optimization (ASO) → delegate to the app-store-optimization skill.
• Build upload / TestFlight / asc release flow / submission → delegate to app-store-release.
• App Review compliance / privacy-manifest deep audit → delegate to app-review-preflight.
• Web/SPA attribution → use a web-analytics skill.
• Android-only attribution / Privacy Sandbox → Android parity notes only.
• Writing the StoreKit/RevenueCat purchase engine → this skill covers pricing strategy and price display; the transaction code (Transaction listener, entitlement gating) belongs to a StoreKit implementation skill.

---

References

File	Covers
references/attribution-and-att.md	ATT timing, 4-state consent gating, SKAdNetwork v4 + AdAttributionKit conversion values (incl. iOS 18 PostbackUpdate, AAK server-side postback receiver, SKAN→AAK dual-call regression), per-channel measurement, MMP troubleshooting, pre-launch checklist
references/tiktok-channel.md	TikTok Business SDK (vs OpenSDK), SKAN-ownership decision matrix, init order, tiktokAppId/appId, hybrid event_id dedup, Events API S2S
references/meta-attribution.md	Meta AEM 7-prereq checklist, "AEM greyed-out" causes, ATE 5-cause list, CUID timing, Conversions API extinfo 16-position positional array
references/apple-search-ads.md	ASA readiness gate, CPT-by-category, Target CPT = CPI × CVR, bid-signal table, CPP routing (+15-30%), KPI ranges, budget tiers, 2025-26 product changes
references/in-app-events.md	In-App Event card specs/limits, indexed-search names, no-opt-in re-engagement push, algorithmic eligibility
references/app-clips-growth.md	App Clips as a discovery→install funnel (3-5×), 15 MB cap, SKOverlay-after-value handoff, card limits
references/metrics-benchmarks.md	Per-category Poor/Avg/Good/Excellent bands, formulas, cohort analysis (3 types + 7 mistakes), sourced 2026 subscription benchmarks
references/ratings-and-reviews.md	requestReview 3/365 limit, when-to-ask table, pre-prompt survey filter (+0.3-0.8★), per-version reset tradeoff, recovery campaign
references/paid-acquisition.md	Platform selection, campaign structure, ad copy/creative, audience targeting, optimization, retargeting, blended CAC
references/monetization-and-pricing.md	Value-based pricing, Van Westendorp/MaxDiff, value metrics, tier structure, freemium vs trial, app unit economics, price increases
references/paywall-pricing-display.md	StoreKit/RevenueCat as price source of truth, locale formatting, derived savings/trial labels, banned Swift literals
references/ppp-territory-pricing.md	asc CLI subscription + IAP territory pricing via CSV import, PPP strategy patterns
references/onboarding-and-cro.md	14-screen questionnaire onboarding framework, permission priming, paywall placement, CRO levers
references/launch-strategy.md	ORB channel framework, 5-phase launch, Product Hunt, ongoing launches
references/referral-and-viral.md	Referral loop design, incentive sizing, K-factor modeling, fraud prevention, tools
references/marketing-psychology.md	70+ mental models mapped to app marketing applications, pricing psychology
references/growth-tactics-library.md	140 growth tactics by category / stage / budget / timeline
references/competitor-pages.md	4 comparison-page formats, centralized competitor data, SEO + schema

Scripts

• scripts/diagnose-att.sh [project-root] — static-analysis pass for iOS marketing + ATT health (CI-friendly, read-only, no network). Flags the init-order no-op (wait config after start()), the capture-protection silent-defer ATT bug, ungated marketing-event call sites, and inventories AppsFlyer/Adjust/Branch/Singular/Meta/TikTok/Firebase/AdMob. Run it as the mechanical first pass of an "is our attribution healthy?" audit, then read references/attribution-and-att.md for the fixes.