How We Increased ROAS by One Third for an E-commerce Fashion Brand Through Proper Ad Segmentation

Key problems before optimization

• campaigns overlapped by audience — warm and cold traffic were mixed together
• the algorithm received mixed signals and could not optimize efficiently
• budget was distributed unevenly across segments with different purchase intent
• ROAS was unstable — scaling immediately reduced profitability

The issue was not the budget or the creatives — it was the structure, which did not give the algorithm clear signals for optimization.

Audience segmentation

We clearly separated three levels:

• cold traffic — new users who had not interacted with the brand yet
• warm traffic — product views and add-to-carts
• remarketing — users with high purchase intent

We removed audience overlap between campaigns — each segment got its own logic and budget.

New Google Ads structure

• separate Search campaigns for each segment
• separate Performance Max campaigns with clear audience signals
• clear budget allocation
• traffic distribution control across segments

Optimization for conversion value

We shifted the focus from the number of conversions to their value. We optimized bids and adjusted audiences based on real order profitability. The algorithm finally got clean signals, which improved ROAS and made scaling more stable.

Result

• $1,250 ad budget
• $11,500 conversion value
• ROAS growth of +33%