3.2x ROAS Improvement for E-commerce Brand
How we restructured a Google Ads account and fixed tracking to achieve 3.2x ROAS improvement in 90 days.
3.2x improvement
ROAS
42% reduction
Wasted Spend
98% accuracy
Conversion Tracking
The Challenge
A mid-sized e-commerce brand came to us with a Google Ads account that was “working but not scaling.” Monthly spend was around €25,000 but ROAS had plateaued at 2.1x despite multiple attempts to optimize.
After our initial audit, we identified three core problems:
- Account structure chaos – Campaigns had grown organically without strategy, resulting in competing keywords and cannibalized budget
- Broken conversion tracking – Enhanced conversions weren’t implemented, leading to 30-40% data loss post-iOS 14
- No negative keyword management – Search terms reports showed significant spend on irrelevant queries
Our Approach
Phase 1: Tracking Foundation (Week 1-2)
Before touching the ads, we fixed the data:
- Implemented enhanced conversions for Google Ads
- Set up server-side conversion tracking via GTM
- Fixed GA4 e-commerce event configuration
- Validated all data with real purchases
Phase 2: Account Restructure (Week 3-4)
We rebuilt the account from scratch:
- Created a clear campaign hierarchy by product category
- Implemented single keyword ad groups for top performers
- Built comprehensive negative keyword lists
- Set up proper Performance Max asset groups
Phase 3: Optimization (Week 5-12)
With clean structure and accurate data:
- Ran systematic bid strategy tests
- Optimized shopping feed attributes
- A/B tested ad copy across campaigns
- Implemented audience layering
Results
After 90 days:
- ROAS improved from 2.1x to 6.7x (3.2x improvement)
- Wasted spend reduced by 42% through better targeting
- Conversion tracking accuracy reached 98% with server-side setup
The client was able to scale spend by 60% while maintaining the improved ROAS.
Key Takeaways
- Fix tracking first – You can’t optimize what you can’t measure
- Structure matters – Clean account architecture enables algorithmic learning
- Negative keywords compound – Regular search term mining prevents budget waste
Want similar results? Book a free audit and we’ll show you what’s possible with your account.