Guide

Game user acquisition and marketing explained

Harbor Puzzle shipped a polished match-three with a fair monetization model and strong day-one retention in soft launch. Global launch week brought 400 organic installs and a celebratory team Slack thread. Then silence. Without paid user acquisition (UA), the game never reached the audience density needed for leaderboards, live events, or ad mediation fill rates to work. Marketing is not a trailer dropped on launch day — it is a measurable funnel from first impression to paying player, tuned until lifetime value (LTV) exceeds cost per install (CPI). This guide explains how game UA differs from generic app marketing, the metrics that matter, creative and channel strategy, store optimization, a Harbor Puzzle mobile launch worked example, a channel decision table, common pitfalls, and a production checklist. It complements game analytics and live ops without replacing either.

UA, marketing, and organic growth: what each owns

Studios blur these terms; budgets suffer when roles are unclear:

User acquisition (UA) — performance marketing with direct spend: Meta, Google UAC, TikTok, Apple Search Ads, rewarded play networks. Success is measured in installs, CPI, and return on ad spend (ROAS) at day 7, day 30, or day 180 depending on genre.
Brand and community marketing — trailers, press kits, Discord, creator relations, convention booths. Harder to attribute per install but builds wishlists, word-of-mouth, and creative assets that paid UA later recycles.
Organic discovery — App Store and Google Play search ranking, featuring, SEO for web games, Steam discovery queues. Optimized through ASO (app store optimization): keywords, screenshots, icon tests, review velocity.
Product-led growth — invites, co-op links, shareable replays, referral rewards baked into the game. Cheap at scale when the core loop is inherently social.

A sustainable F2P title usually combines all four. Premium PC games lean heavier on wishlists and influencers; hyper-casual mobile lives almost entirely on paid UA until creative fatigue kills CPI.

The install funnel and metrics that matter

Think in stages; optimize the weakest link before scaling spend:

Top of funnel: impression to install

IPM (installs per mille) — installs per 1,000 ad impressions. Combines CTR and store conversion; the creative team’s headline KPI on many networks.
CPI (cost per install) — ad spend divided by attributed installs. Varies wildly by geo, platform, and genre ($0.20 hyper-casual India vs $8+ US casino).
Store conversion rate (CVR) — product page views to installs. Low CVR with high IPM means the ad oversells; fix screenshots and first-screen copy before blaming the game.

Middle of funnel: install to engaged user

FTUE completion rate — % reaching level 3, first multiplayer match, or whatever marks “got it.” Tied directly to tutorial design.
D1 / D7 / D30 retention — cohort curves from analytics. UA that buys low-quality installs shows up here as cliff-drop D1, not just bad CPI.
Session length and sessions per day — engagement depth; needed to model ad revenue and IAP timing.

Bottom of funnel: engagement to revenue

ARPDAU — average revenue per daily active user (IAP + ads).
LTV (lifetime value) — predicted revenue per install over 30, 90, or 365 days. UA scales only when LTV > CPI (plus platform fees and attribution lag).
ROAS — revenue attributed to ad spend divided by that spend. ROAS 100% at D30 means break-even on a 30-day horizon; studios target 120–150%+ before scaling aggressively.
Payer conversion and ARPPU — what fraction spend and how much whales contribute; see economy design.

Run UA math on cohorts, not lifetime totals. A spike in installs without cohort tagging makes ROAS look worse than it is for weeks.

Paid UA channels and when to use them

Social and video networks

Meta (Facebook/Instagram), TikTok, and Google App Campaigns dominate mobile UA. They reward creative volume: dozens of short video variants tested weekly. Gameplay capture, fail-state hooks (“only 1% pass level 47”), and UGC-style vertical video often beat cinematic trailers. Broad targeting plus strong creatives frequently beats hyper-narrow lookalikes early in a campaign.

Apple Search Ads and Google Search

High-intent users already searching your genre or competitor names. Lower volume, often lower CPI volatility. Essential for branded defense once copycat apps bid on your title.

Rewarded play and offer walls

Users install to earn currency in another game. CPI can look attractive; retention often weaker. Useful for volume tests or secondary geos, risky as primary channel for mid-core titles.

PC and console: Steam, Epic, consoles

Wishlist velocity drives algorithmic visibility. Demo festivals (Steam Next Fest), influencer keys, and Metacritic-sensitive press matter more than mobile-style CPI. Track wishlists-to-sales conversion at launch.

Influencers and creator campaigns

Sponsored streams and YouTube integrations. Negotiate whitelisting rights so their best-performing clips become paid UA creatives. Measure with custom links and promo codes, not vibes.

Cross-promotion and ad mediation

If you operate multiple titles, house ads in your existing player base are the cheapest installs you will ever get. Mediation waterfalls can include cross-promo inventory before third-party networks.

Creative testing: the real UA product

On most networks, creative is the targeting. A single hit video can cut CPI in half; fatigue can double it in three weeks. Operational habits that work:

Produce 5–10 new concepts per week from one gameplay build — different hooks, openings, end cards, and CTA text.
Test hooks in the first 2 seconds: question, fail, reward, or curiosity gap. Mobile users scroll faster than your logo animation.
Localize captions and VO for tier-1 geos; same gameplay, different language overlays.
Kill losers at $50–$200 spend; promote winners to dedicated campaigns before they fatigue.
Align ad promise with core loop reality. Misleading ads boost IPM and destroy D1 retention — negative ROAS even at cheap CPI.

Store assets are creatives too. A/B test icons and screenshot order in Google Play Experiments and Apple Product Page Optimization. The winning icon often differs from the winning ad thumbnail.

Attribution, privacy, and SKAN

iOS App Tracking Transparency (ATT) and SKAdNetwork limit granular user-level attribution. Android privacy sandboxes are moving similarly. Practical implications:

Modeled and probabilistic attribution fill gaps; treat as directional, not gospel.
SKAN postbacks arrive delayed and aggregated; day-7 ROAS dashboards lag iOS reality by days.
Incrementality tests (geo holdouts, pause tests) validate whether paid installs are incremental or would have happened organically.
First-party data — account login, email, server-side purchase events — becomes more valuable for LTV modeling as IDs disappear.

Document your MMP (Adjust, AppsFlyer, Singular, etc.) event schema before launch. Retrofitting purchase events after UA spend starts wastes learning weeks.

Worked example: Harbor Puzzle global launch

After soft launch in Philippines and Canada, Harbor Puzzle had D1 retention 42%, D7 18%, and ARPDAU $0.06 blended (ads + light IAP). The team modeled 90-day LTV at roughly $0.45 in tier-1 geos and $0.22 in tier-3.

Launch plan:

ASO pass — keyword research on competitor titles, new screenshot set showing combo cascades, localized store listings for EN, JP, DE. Organic installs rose 30% before any paid spend.
Creative sprint — 24 vertical videos in two weeks: fail-state hooks, satisfying chain reactions, “no timer” calm positioning vs hyper-casual competitors. Top three creatives averaged IPM 11 vs portfolio baseline 6.
Geo tiering — US/UK/AU campaigns capped at CPI $2.50 until D7 ROAS model confirmed; scale paused when blended CPI hit $3.10 with D1 retention unchanged (creative fatigue signal).
Influencer whitelisting — three mid-tier puzzle YouTubers; best clip reused as paid creative, cutting CPI 22% on Meta.
Cross-promo — house ads in Harbor Arcade drove 8% of week-one installs at near-zero marginal cost.

After six weeks, tier-1 campaigns stabilized at CPI $2.20, D30 ROAS 118%. Scaling continued into Season 1 live ops only after LTV curves confirmed — not on launch-week excitement.

Channel decision table

Studio situation	Prioritize	Deprioritize until…
Hyper-casual mobile, short sessions	Meta/TikTok creative volume, rewarded tests	Influencer long-form (poor fit)
Mid-core F2P, strong D7 retention	Social video + ASA, LTV-based bid caps	Blind scale before D30 LTV modeled
Premium PC indie	Steam wishlists, demos, streamers	Mobile-style CPI campaigns
Portfolio with existing DAU	Cross-promo, email/push to sister titles	Buying cold traffic before house ads saturated
Pre-launch, limited budget	ASO, community Discord, vertical slice playtests	Global paid blast on unproven FTUE
Live service with seasons	Retargeting lapsed players, creative tied to season hook	Only prospecting without win-back layer

Common pitfalls

Scaling before LTV is modeled — cheap CPI with terrible D7 retention burns cash faster than expensive CPI with loyal players.
One trailer, zero variants — networks need creative churn; a launch trailer is not a UA strategy.
Misleading ads — spikes installs, craters retention, triggers store policy risk.
Ignoring store CVR — optimizing ads while product page screenshots show menu screens nobody clicks through.
No cohort tagging — impossible to compare channel quality or run A/B tests on landing experiences.
UA without live ops capacity — buying players into a static game is a leaky bucket; align spend with content cadence.
Attribution worship — last-click MMP numbers without incrementality tests overcredit paid and underfund organic.

Production checklist

Define target geos, platforms, and CPI/LTV thresholds before first dollar spent.
Instrument MMP events: install, FTUE milestones, purchase, ad impression, session start.
Build 10+ ad-ready video variants from real gameplay, not only cinematics.
Localize store listings and top creatives for priority markets.
Run soft launch in 1–2 geos; validate retention and ARPDAU before tier-1 scale.
Set weekly creative production quota and kill/scale rules by spend tier.
Align UA hooks with actual onboarding and core loop — no bait-and-switch.
Plan cross-promo and influencer assets before global launch day.
Review cohort ROAS at D7, D30, and D90; adjust bids and geos on curves, not vanity CPI.
Pair acquisition spikes with live ops beats so new players meet timely content.

Key takeaways

UA is a funnel, not a launch event — impression, install, engage, monetize, each with its own KPI.
LTV must exceed CPI on cohort horizons that match your payback target — otherwise scale makes losses scale too.
Creative testing is continuous — the best UA teams behave like content studios.
Organic and paid reinforce each other — ASO, community, and cross-promo lower blended acquisition cost.
Product quality is the ceiling — marketing finds players; retention and monetization determine whether they were worth finding.