Google App Campaigns Mastery: UAC Optimization Guide

Google App Campaigns is a black box. You throw in some text, images, videos, and a budget, and Google's machine learning does the rest. For years, I hated this lack of control. Then I learned to love it.

My education began when I was managing UA for a puzzle game. Our Meta campaigns were crushing it, but we'd hit a ceiling. A colleague suggested we get serious about Google. I was skeptical—UAC felt like shouting into a void. But we committed, learned the system, and within six months, Google became our largest and most profitable channel.

This guide is everything I learned on that journey.

Understanding Google App Campaigns

Google App Campaigns (formerly Universal App Campaigns or UAC) is Google's automated campaign type for driving app installs and in-app actions. It runs across Google's entire inventory:

• Google Search: App install ads in search results
• Google Play: Ads in the Play Store
• YouTube: In-stream and discovery ads
• Display Network: Banner ads across millions of sites
• Discover: Native content feed ads
• Gmail: Promotions tab ads
• AdMob: In-app ads within other apps

Unlike Meta where you can control targeting precisely, Google's algorithm makes all placement and audience decisions. Your job is to provide quality inputs and let the machine optimize.

Campaign Types and Objectives

Google offers three main optimization goals for app campaigns:

App Installs (tCPI)

Optimizes for maximum install volume at your target cost-per-install. Best for:

• Early-stage apps building user bases
• Apps with strong organic monetization
• Broad appeal apps where volume matters

In-App Actions (tCPA)

Optimizes for users who complete specific events after install. Best for:

• Apps with defined conversion funnels
• When you need quality over quantity
• Subscription and e-commerce apps

In-App Action Value (tROAS)

Optimizes for revenue, seeking users who will spend the most. Best for:

• Mature apps with purchase history
• Gaming apps with IAP revenue
• Apps where LTV varies dramatically

Asset Strategy: Feeding the Machine

Google's algorithm creates ads by combining your assets. More quality assets = better performance. Here's what you need:

Text Assets

• Headlines: 2-5 headlines, up to 30 characters each
• Descriptions: 1-5 descriptions, up to 90 characters each

Best practices:

• Lead with benefits, not features
• Include your app's unique value proposition
• Test different angles (urgency, social proof, curiosity)
• Keep some headlines short for smaller placements

Image Assets

• Minimum: 1 landscape (1200×628), 1 portrait (300×250)
• Recommended: 20 images across all sizes
• Sizes: 320×50, 320×100, 300×250, 320×480, 1200×628

Image tips:

• Show the app in action, not just icons
• Test screenshots vs lifestyle imagery
• Include clear CTAs in images
• Avoid too much text (20% rule applies)

Video Assets

Video is crucial for YouTube and high-impact placements:

• Minimum: 1 video
• Recommended: 5-10 videos
• Orientations: Landscape (16:9), Portrait (9:16), Square (1:1)
• Lengths: 10-30 seconds for best performance

HTML5 Playable Assets

For gaming apps, playable demos can dramatically improve conversion:

• Interactive mini-games that showcase core gameplay
• 15-60 second experiences
• Higher production cost but often highest performance

Bidding Strategy Deep Dive

Your bid tells Google how much you're willing to pay—and influences which users you reach.

Setting Target Bids

Start conservative, then adjust based on performance:

1. Calculate your maximum allowable CPI/CPA based on LTV
2. Set initial target at 80% of maximum
3. Let the campaign run for 2-3 weeks
4. Adjust based on actual performance and volume needs

The Bid-Volume Trade-off

Higher bids = more volume but higher costs. Lower bids = better efficiency but less scale. Find your sweet spot:

• Volume Phase: Bid higher to acquire data and train the algorithm
• Efficiency Phase: Lower bids to optimize margins
• Scale Phase: Find the bid that maximizes profitable volume

Geo-Specific Bidding

Unlike Meta, Google doesn't let you bid differently by country within a campaign. Solution: Create separate campaigns for different geo tiers:

• Tier 1: US, UK, CA, AU—higher bids, higher LTV
• Tier 2: Western Europe, Japan, Korea—moderate bids
• Tier 3: LATAM, SEA, Eastern Europe—lower bids
• Tier 4: ROW—lowest bids or exclude

Learning Phase and Optimization

Google's algorithm needs time to learn. Here's how to navigate it:

The Learning Phase

New campaigns enter a learning phase where performance is volatile:

• Duration: Typically 2-4 weeks
• Required: 10-15 conversions per day minimum
• Behavior: CPIs/CPAs may exceed targets initially

What Triggers Learning Reset

• Bid changes greater than 20%
• Budget changes greater than 30%
• Changing optimization goal
• Adding/removing significant assets

Optimization Levers

Once stable, your optimization options are:

• Asset refresh: Replace underperforming assets
• Bid adjustment: Small incremental changes (10-15%)
• Budget scaling: Gradual increases (20% max)
• New campaigns: Test different goals or geos

Advanced Strategies

Campaign Segmentation

Don't put all eggs in one campaign. Segment by:

• Geography: Separate tiers for different bid levels
• Objective: Install campaigns vs action campaigns
• Creative theme: Test different messaging approaches
• Language: Localized assets for non-English markets

Pre-Registration Campaigns

For new game launches, Google offers pre-registration campaigns:

• Build anticipation before launch
• Users get notified when the app releases
• Lower cost than post-launch acquisition
• Creates launch momentum

Custom Store Listings

Create custom Play Store pages for different audiences:

• Different screenshots for different user segments
• Localized descriptions and imagery
• Link specific campaigns to specific store listings
• Improve post-click conversion rates

Measurement and Attribution

Google Ads Conversion Tracking

Google's native conversion tracking is required for campaign optimization but has limitations:

• Only sees Google-driven conversions
• Can conflict with MMP data
• Uses its own attribution model

MMP Integration

Your MMP (AppsFlyer, Adjust, etc.) should be your source of truth. Integrate properly:

• Link Google Ads to your MMP
• Pass conversion events to Google for optimization
• Use MMP for cross-channel attribution
• Reconcile data differences monthly

View-Through Attribution

Google counts engaged-view conversions (users who watched 10+ seconds of a video). This can inflate numbers compared to click-only attribution. Understand and account for this difference.

Common Google App Campaign Mistakes

• Too few assets: 3 headlines and 2 images starves the algorithm
• Impatience: Killing campaigns before learning completes
• Over-optimization: Constant changes prevent learning
• Ignoring video: Missing YouTube inventory opportunity
• Generic creative: Stock photos and bland copy
• Wrong objective: Using tCPI when you should use tCPA
• Budget constraints: Daily budget too low for proper learning

Google vs. Meta: When to Use Each

Both platforms have strengths. Here's how to think about allocation:

Google Strengths

• Massive Android reach (especially emerging markets)
• Intent-driven placements (Search, Play Store)
• Often lower CPIs, especially for casual games
• YouTube's engaged video audience

Meta Strengths

• Stronger iOS performance (especially post-ATT)
• Better creative control and testing
• Superior user quality signals
• More granular optimization options

Allocation Framework

A common starting point:

• Android: 50-60% Google, 40-50% Meta
• iOS: 30-40% Google, 60-70% Meta

Adjust based on your specific performance data.

That puzzle game I mentioned? We eventually reached a 60/40 Google/Meta split on Android and generated 40% of our total revenue from Google-acquired users. The black box works when you feed it right.