Google Ads Case Study

Scaling an E-commerce Account 60%+ While Holding ROAS Above Target

Specialist automotive parts retailer  ·  Performance Max + Shopping + Search  ·  90-day period

The client

A specialist automotive parts e-commerce retailer running a lean Google Ads account across three live channels: a feed-based Performance Max campaign for its Jeep product line, a non-brand Shopping campaign for its wider catalogue, and a brand Search campaign. The account carried a hard efficiency mandate — a 3.5x return-on-ad-spend (ROAS) target that revenue growth could not be allowed to erode.

The challenge

The account was profitable but carrying structural inefficiencies that capped its ceiling, and it needed to scale spend without slipping below target. Three issues stood out:

• Match-type leakage in the brand campaign — broad-match brand terms were converting at roughly 2.4x ROAS while the same terms on exact match returned about 3.7x. Spend was leaking into a weaker path.
• A silent conversion-path failure — the Jeep Performance Max feed stopped recording conversions for roughly ten days, the kind of drop that is easy to misdiagnose as an ad-platform problem and “fix” in the wrong place.
• Scaling pressure without headroom — the goal was more revenue and volume, but every efficiency-losing move was off the table given the 3.5x floor.

The approach

I ran the account on a disciplined weekly review cadence — reading every metric against both a 30-day baseline and a 7-day recent-movement window, so nothing got called a trend on noise. The work concentrated on structure and diagnosis rather than churn:

• Tightened match-type discipline — paused the underperforming broad brand keywords, migrated the salvageable term to phrase match, and expanded exact-match coverage so brand spend flowed to its most efficient path.
• Diagnosed the conversion drop correctly — segmented Performance Max by feed vs. non-feed performance to isolate the failure to the Shopping-inventory path, then traced the root cause to a store-side issue rather than the ad account, avoiding days of wasted troubleshooting in the wrong system.
• Protected the structure that was working — resisted reactivating paused non-brand Search campaigns that would have cannibalised volume already converting efficiently through Shopping, and kept Performance Max delivery attributed to the correct lever rather than throwing budget at it.
• Held ruthless search-term and negative-keyword hygiene — flagged genuinely wasteful spend while deliberately not negativing low-volume terms that carried no real signal, keeping the account clean without starving it of data.

The results

Over the latest 90-day period, ad spend was scaled roughly two-thirds while revenue and purchase volume grew in step and ROAS held above the 3.5x target — efficient scaling rather than growth bought at the cost of margin.

Why it worked

The gains came from diagnosis and discipline, not spend for its own sake: fixing where money leaked, reading platform signals correctly before acting, and leaving the parts of the account that were already performing well alone. That combination is what let the account grow revenue and conversions in the 60%+ range while keeping return on ad spend above target.