Case Study 1

Case Study / Google Ads Rebuild

The fix wasn't a bigger budget. It was a three-campaign structure, each doing one job.

A direct-to-consumer Jeep accessories brand inherited an underperforming Google Ads account: minimal spend, three conversions a month, no clear structure. Thirty days after a full rebuild, the same account returned 16x on ad spend across 156 conversions, on a budget under £1,300.

Spend $1,205 +830%

Revenue $19,308 +3,460%

Conversions 156 +5,100%

ROAS 16.01x +283%

01 / The Brief

An ecom account running cold, with no clear job for any campaign.

The brand sells niche Jeep accessories direct to consumer. Strong product, strong margin, loyal community on the organic side. The Google Ads account did not reflect any of that.

When the account came across, it had been sitting in a holding pattern for months: roughly $130 a month in spend, three conversions to show for it, and a structure that made it almost impossible to read what was actually happening.

Brand search, generic prospecting, and Performance Max were all running through a single tangled structure with overlapping keywords, shared budgets, and no clear separation of intent. Brand traffic was being attributed to PMax. PMax was eating cheap branded clicks and reporting them as prospecting wins. Generic search was getting throttled because there was no dedicated budget protecting it.

The first job was diagnostic, not creative. Before touching a single ad, the priority was working out what each campaign was supposed to do, and rebuilding the account so the answer to that question was unambiguous.

02 / The Reset

30 days before, and 30 days after.

Same account. Same products. Same audience. The only thing that changed was structure, tracking, and how budget was allocated against intent.

Before / Feb 25 – Mar 26 Spend $129.68 Revenue $542.43 Conversions 3 ROAS 4.18x

After / Mar 27 – Apr 25 Spend $1,205.75 Revenue $19,308.31 Conversions 156 ROAS 16.01x

03 / The Rebuild

Three campaigns. Three jobs. No overlap.

The principle was simple: every campaign should have one job, one budget, and one way of being judged. If a campaign could not answer the question "what are you here to do," it did not belong in the account.

Campaign 01 Brand Search Defend branded queries. Capture intent already in the funnel. Stop PMax from claiming credit for traffic it did not earn. Job: Protect existing demand.

Campaign 02 Generic Search Tightly themed ad groups around high-intent product categories. Branded terms excluded as negatives. Manual CPC to start, tCPA once the conversion data was clean. Job: Win new buyers actively searching.

Campaign 03 Performance Max Brand terms excluded at the account level. Audience signals fed in. Asset group rebuilt with full creative library, lifestyle imagery, and video. Budget capped to prevent it eating the account. Job: Scale incremental, non-brand demand.

Tracking was rebuilt at the same time. Purchase value passed back cleanly, brand and non-brand reporting separated, and a simple weekly review cadence put in place so optimisation decisions were based on the right denominator, not the platform's default view.

04 / What Moved the Needle

Structure first. Creative second. Budget last.

01	Separating brand from non-brand alone exposed where the real waste was. Roughly 40% of pre-rebuild conversions were branded clicks being claimed by PMax.
02	Asset density in PMax moved the algorithm from guessing to learning. More headlines, more lifestyle imagery, audience signals, and a proper product feed gave the system enough surface area to optimise against.
03	Negative keyword discipline in generic search cut off the wasted clicks that had been masking what the account could actually do. Search term reviews twice a week for the first 14 days.
04	Budget was the last lever pulled, not the first. Spend went up because the structure could finally absorb it profitably. If the rebuild had not come first, more budget would have meant more waste at scale.