← All Work

02 -Paid Social & User Acquisition

User Acquisition
Case Study.

How a Bartle-inspired persona × motivation framework replaced demographic-only targeting -and drove CPI to $0.21 with D7 ROAS of 13.5%.

The Problem

Demographic targeting (35F, AMER) tells you nothing about why someone plays. Two women, same profile — one tops leaderboards, another decorates her restaurant. Same ad, different person. Doesn't work.

The Approach

Bartle taxonomy: classify players by motivation, not demographics. 4 personas, 3 motivations each, 12 creative briefs — each targeting a specific emotional trigger.

My Role

UA Strategy Persona Development Creative Framework A/B Testing Campaign Management

Target: 35+ women, AMER & EU

Slide 1 of 2 -Persona Mapping

4 Personas -Where They Sit on the Bartle Quadrant

Axes: how a player interacts with the game world (left) vs. other players (right), and whether they act (top) or interact (bottom)

Acting Interacting World Players
🏆
Achiever
⚔️
Killer
🔍
Explorer
🦋
Social Butterfly
Achiever
Motivated by advancement -completes levels, earns stars, unlocks new restaurant tiers. Progress is the reward.
Killer
Competitive and ranking-driven. Plays to win, tops leaderboards, enters tournaments. Needs a rival to beat.
Explorer
Discovery-driven. Loves new ingredients, equipment unlocks, and decoration options. Plays to find new things.
Social Butterfly
Community-driven. Gifting, co-op cooking events, guild life. The game is a social space first, a game second.

4 Personas × 3 Motivations = 12 Creative Buckets

Each motivation = a distinct creative brief with its own hook, narrative, and CTA

Achiever
Progress & mastery
Level Completion
Show the satisfaction of 3-star clears and progression milestones
Restaurant Expansion
Unlocking new tiers, bigger kitchens, new cuisines
Skill Mastery
Master the core cooking loop -speed, accuracy, combos
Killer
Competitive & ranking
Tournament Rankings
Seasonal competitions, climbing the leaderboard
Restaurant vs. Restaurant
Head-to-head battles, beating rival restaurants
Seasonal Events
Limited-time competitive events with exclusive rewards
Explorer
Discovery & variety
New Ingredients & Unlocks
Rare ingredients, exotic cuisines, hidden game content
Decoration & Design
Customise the restaurant -furniture, themes, layouts
Equipment Upgrades
New kitchen equipment, expanded cooking capabilities
Social Butterfly
Community & connection
Gifting & Helping
Sending gifts to guild members, helping friends pass levels
Co-op Cooking Events
Collaborative team challenges and community cooking events
Guild Presence
Visibility in the community, guild leadership, social recognition

4 personas × 3 motivations = 12 creative briefs. Each brief produced 2–3 ad variants for A/B testing.

The Operational Model

12 briefs, 2-3 variants each, running against segmented audiences simultaneously. Metrics tracked per bucket: CPI, CTR, D7 ROAS. Winners scaled, losers cut — but every loss logged a reason. Did the creative fail, or did the targeting miss?

Why It Worked

Every creative had a hypothesis. Random testing gives you winners. Systematic testing gives you winners plus the reason why — which makes the next cycle faster and cheaper.

The Testing Loop


1
Define the Brief
Each persona x motivation cell = one brief: target emotion, gameplay hook, visual tone, CTA. 12 per cycle.
PersonaMotivationHookCTA
2
Produce Ad Variants
2-3 variants per brief. Same emotional hook, different visual execution. Static, video, or playable depending on channel.
StaticVideoPlayableIn-house production
3
Deploy with Segmented Targeting
Meta Ads + Google UAC. Segmented by region (AMER/EU) and age (35+F). Each bucket runs in isolation for clean data.
Meta AdsGoogle UAC35+F · AMER/EU
4
Measure
CPI, CTR, D7 ROAS per bucket. Install quality signals cross-referenced against persona-motivation combination.
CPICTRD7 ROASD7 RetentionIn-app events
5
Scale Winners / Kill Losers
Low CPI + healthy D7 ROAS = scale. Underperformers paused, reason logged. Creative failure or targeting miss — matters which.
Budget reallocationCreative fatigue tracking
6
Feed Learnings Back
Learnings update the next cycle's briefs. The matrix evolves with the data.
Brief refinementAudience insightCreative evolution
↺ Each cycle produced sharper briefs and tighter targeting. The $0.21 CPI was the output of iterated learning, not a one-off.

Performance Metrics

$0.21
Lowest CPI achieved
Industry average for casual mobile games: $1–3. Motivation-matched creatives consistently outperformed demographic-only campaigns.
13.5%
D7 ROAS (peak)
By Day 7, 13.5% of ad spend recovered through in-app revenue. Full payback in mobile gaming takes months. D7 is the early quality signal. Industry benchmark: 10-20%.
12
Creative buckets tested
4 personas × 3 motivations. Each bucket ran independently, generating persona-level insight on what drives install quality.

Ad Creative Samples

Representative creatives from the persona x motivation matrix. Click to watch on YouTube Shorts.

Ad Creative Sample 1

Ad Sample 1 - Watch on YouTube Shorts ↗
Ad Creative Sample 2

Ad Sample 2 - Watch on YouTube Shorts ↗
← Brand Marketing Next: SEO →