How many creatives should you test at once?

A structured testing approach typically runs 5-10 new creative concepts per week in small test budgets ($50-$200 per creative). This provides enough statistical data to identify winners within 3-5 days while keeping testing costs manageable. Scale winners with larger budgets once validated.

How long should you run a creative test?

Most creative tests need 3-7 days and at least 10,000 impressions per creative to generate statistically meaningful results. Shorter tests may miss slower-converting audiences, while longer tests waste budget on obvious underperformers. Kill creatives that show no promise after 3 days.

What metrics should you use to evaluate creative performance?

IPM (Installs Per Mille) is the primary metric for creative evaluation because it isolates creative quality from bidding dynamics. Secondary metrics include CTR, CVR (app store conversion rate), and early retention signals like Day 1 retention. For mature campaigns, ROAS is the ultimate success metric.

Should you test one variable at a time or multiple variables?

For concept testing, test distinct creative concepts (multiple variables) to find the best direction. For optimization of a winning concept, test one variable at a time (color, copy, character, hook) to understand what drives performance. This two-phase approach balances exploration speed with optimization rigor.

How do you scale a winning creative without causing fatigue?

When a creative tests well, immediately produce 10-20 variations that modify secondary elements while keeping the core winning concept intact. Launch these variations on a staggered schedule across different audience segments and networks to extend the creative's total lifespan.

UA Creative Testing | How to Test Ad Creatives for Mobile Game UA

UA creative testing is the systematic process of developing, launching, and evaluating multiple ad creative concepts to identify the most effective ones for mobile game user acquisition campaigns. It is the discipline that separates high-performing UA teams from those that waste budget on underperforming ads, and it has become the single most important capability for mobile game growth.

Why Creative Testing Is the Highest-Leverage UA Activity

In mobile game user acquisition, creative is the primary variable that determines campaign success. Targeting algorithms on major ad networks like Meta, Google, TikTok, and Unity have become highly commoditized. Every advertiser has access to the same machine learning-driven audience optimization. The differentiation comes from the creative.

Data from leading mobile game studios consistently shows that:

The top 10% of creatives generate 80%+ of total installs in a portfolio.
Creative quality accounts for 50-70% of the variance in CPI performance.
Studios that test more creative concepts per month achieve 30-50% lower CPIs than those with low testing velocity.

For UA managers, this means that the creative testing pipeline is not a support function — it is the core growth engine.

The Creative Testing Framework

A structured creative testing methodology has four phases: ideation, production, testing, and iteration.

Phase 1: Ideation

Creative ideation starts with understanding what has worked before and generating hypotheses about what might work next:

Competitive analysis: Study ads from competitors and top-performing games in your genre. Ad intelligence tools show which creatives are running the longest (a proxy for performance).
Concept mapping: Categorize creative approaches by hook type (gameplay, story, humor, challenge, transformation), visual style, and format.
Data-driven hypotheses: Review your historical performance data to identify patterns. Which hooks, visual styles, and formats generated the highest IPM?
Trend awareness: Monitor cultural trends, meme formats, and emerging creative styles that could translate to ad concepts.

The goal of ideation is to generate a pipeline of 20-50 creative concepts per month that span diverse approaches. Relying on a narrow range of concepts accelerates creative fatigue across your entire portfolio.

Phase 2: Production

Production is traditionally the biggest bottleneck in creative testing. Designers can only produce a limited number of polished concepts per week, which constrains testing velocity.

AI-powered creative tools fundamentally change this dynamic. With Layer, teams can:

Generate dozens of visual variations from a single concept brief using 300+ AI models for image and video.
Apply custom style training to ensure generated assets match the game's visual identity.
Use workflow automation to create production pipelines that output finished ad creatives, not just raw assets.

This shifts the production bottleneck from "can we make enough creatives?" to "can we test enough creatives?" — a much better problem to have.

Phase 3: Testing

Structured testing follows a clear protocol:

Test setup:

Allocate a fixed test budget per creative (typically $100-$300 for initial validation).
Run each creative on the same audience targeting and ad network to ensure comparability.
Use a consistent ad format within each test cohort (do not mix video and static in the same test).
Allow 3-5 days for data collection with a minimum of 10,000 impressions per creative.

Evaluation criteria:

Primary metric: IPM — measures creative effectiveness independent of bidding dynamics.
Secondary metrics: CTR (hook strength), CVR (store conversion alignment), video completion rate (for video ads).
Kill criteria: Creatives that underperform the genre benchmark by 50%+ after 3 days should be paused immediately.
Winner criteria: Creatives that exceed the benchmark by 20%+ move to the scaling phase.

Statistical rigor: Not every test difference is meaningful. Use statistical significance testing (confidence level of 90%+ minimum) before declaring winners and losers. Small sample sizes with apparent performance differences often converge as more data arrives.

Phase 4: Iteration

Iteration is where testing velocity compounds into sustained competitive advantage:

Winner expansion: Take winning concepts and produce 10-20 variations, modifying secondary elements (colors, backgrounds, text, character positioning) while preserving the core hook.
Format translation: Translate winning static concepts into video, and winning video concepts into playable ads.
Audience extension: Test winners on new audience segments and ad networks.
Fatigue monitoring: Track each creative's IPM trajectory and schedule replacements before performance degrades significantly.

Creative Testing at Scale with AI

The economics of creative testing favor volume. Studios that test more concepts find more winners, and more winners mean lower CPI and more efficient growth. Here is how AI tools enable testing at unprecedented scale:

Concept Exploration

AI image generation allows teams to explore visual concepts in minutes rather than days. A single prompt can generate dozens of visual directions, giving creative strategists a broad palette to choose from before committing production resources.

Variation Generation

Once a winning concept is identified, AI tools can generate hundreds of variations automatically. Layer's platform connects image and video AI models to production workflows, enabling teams to produce variation batches on demand. This is critical for combating creative fatigue — the faster you can produce variations of a winner, the longer you can sustain its performance.

Style Consistency

A key concern with AI-generated creatives is maintaining brand consistency. Layer's custom style training solves this by fine-tuning models to match your game's specific visual identity. This means AI-generated variations feel cohesive with your existing creative portfolio while still exploring new visual territory.

Building a Creative Testing Organization

Effective creative testing requires more than tools. It requires organizational alignment:

Dedicated creative strategist: Someone who bridges data analysis and creative ideation. They interpret test results and generate hypotheses for the next round of testing.
Clear metrics dashboard: Real-time visibility into IPM, CPI, and creative lifecycle metrics so the team can react quickly.
Rapid feedback loops: Weekly creative reviews where the team analyzes test results, identifies patterns, and plans the next batch of tests.
Knowledge base: Document winning concepts, failed approaches, and insights so the team builds institutional knowledge over time.

Growth leads who invest in this organizational infrastructure consistently outperform those who treat creative testing as an ad hoc activity. Combined with AI-powered production tools like Layer, a well-structured creative testing operation becomes a sustainable moat that compounds over time.

Measuring Creative Testing ROI

To justify investment in creative testing infrastructure, track these metrics:

Win rate: Percentage of tested concepts that meet the "winner" threshold. Industry average is 10-15%; top teams achieve 20-25%.
CPI trend: Month-over-month CPI trajectory. Healthy testing programs show flat or declining CPIs even as spend scales.
Creative lifespan: Average number of days before a creative fatigues. Effective iteration extends lifespan by 50-100%.
Testing velocity: Number of concepts tested per month. Aim for 50+ to maintain a healthy pipeline.

These metrics demonstrate that creative testing is not a cost center but rather the primary driver of UA efficiency and growth.

UA Creative Testing: How to Build a Winning Testing Methodology