How We Scaled Google Ads ROAS from 1.8x to 5.2x in 6 Months

When we first audited the account, we found what we see in 80% of struggling Google Ads accounts: a bloated structure with hundreds of ad groups, no proper audience segmentation, and Smart Bidding strategies fighting each other for the same conversions.

The Problem: A House of Cards

The brand had been running Google Ads for 3 years. They'd accumulated 847 ad groups across 12 campaigns. Their Shopping campaigns were cannibalising their Search campaigns. Their Performance Max campaigns were eating budget that should have gone to their best-performing Shopping segments.

ROAS was 1.8x. They needed 3x to break even on customer acquisition.

Our Approach: Ruthless Simplification

The first thing we did was controversial: we paused 70% of their campaigns.

Not deleted — paused. We needed to understand what was actually driving conversions before we rebuilt anything.

Step 1: Conversion Tracking Audit

Before touching bids or budgets, we audited every conversion action. We found:

Once we fixed the tracking, their "real" ROAS was actually 2.3x — better than we thought, but still not profitable.

Step 2: Campaign Architecture Rebuild

We rebuilt the account around three principles:

**Principle 1**: One campaign per intent level (awareness, consideration, purchase)

**Principle 2**: Shopping and Search should complement, not compete

**Principle 3**: Performance Max gets the budget it earns, not a fixed allocation

The new structure had 8 campaigns total. Clean. Auditable. Scalable.

Step 3: The Creative Testing Framework

This is where most accounts leave money on the table. We implemented a 3-2-2 testing framework:

Every ad group had at least 6 ads in rotation. We let Google's RSA system do the heavy lifting, but we fed it better inputs.

The Results

After 6 months:

The key insight? Most Google Ads accounts don't need more campaigns. They need fewer, better-structured ones with cleaner data feeding the algorithm.