Case Study: UK Custom Number Plate Retailer

Results

Increased YoY lead flow by 90% and decreased cost per conversion by 54%

Before optimisation, the account was technically generating conversions, but the data was not clean enough to scale confidently.

• Conversion tracking included add to carts alongside purchases
• Conversion volume looked healthy on paper but overstated real performance
• Campaigns were competing against each other in auctions
• Branded traffic was leaking into non branded campaigns

These issues were limiting efficiency and making it harder to understand what was actually driving revenue.

Our Approach

Conversion Tracking and Data Quality

The first priority was fixing conversion quality.

Add to cart actions were removed from primary conversion tracking so the account was optimised purely toward completed purchases. This immediately aligned bidding and optimisation decisions with real revenue.

Enhanced conversions were then enabled, and conversion tracking was rebuilt through Google Tag Manager. This improved attribution accuracy and gave Google stronger signals to work with, particularly for lower funnel users.

Once tracking was cleaned up, performance data became far more reliable.

Shopping Campaign Structure

A fully segmented standard Shopping structure was rolled out, with separate campaigns built around individual product categories.

Each campaign was managed independently, with clear performance targets and tighter control. This allowed spend to flow only into product categories that could maintain a 3+ return, rather than having strong and weak products blended together.

This structure made scaling safer and more predictable.

Reducing Campaign Cannibalisation

Auction overlap was a clear issue in the account. Campaigns were competing against each other for the same search terms, driving up costs without improving results.

To fix this, negative keywords were introduced across campaigns to block terms that were being intentionally targeted elsewhere. This reduced internal competition and helped each campaign do the job it was built for.

Branded Traffic Control

Branded keywords were removed from all non branded campaigns. This ensured branded demand was not inflating performance metrics across the account and allowed non branded campaigns to be judged on their true ability to generate new revenue.

This change also improved visibility into where growth was actually coming from.

Our Results

• ROAS increased to 3.44, up by 1.35 compared to the previous period
• Conversion value increased by £3,665.32, reaching £8,795.43
• Conversions increased by 80, reaching 236 purchase actions
• Clicks increased by 385, reaching 5,876 total
• Impressions decreased by 14,121, reflecting tighter, more relevant targeting

Despite serving fewer impressions, the account generated more clicks, more conversions, and significantly higher revenue. This confirmed that traffic quality improved as tracking and structure were cleaned up.

Why This Worked

Performance improved because the account stopped optimising around noise and started optimising around revenue.

• Google learned from purchases instead of add to carts
• Shopping campaigns were structured for control, not convenience
• Auction overlap was actively reduced
• Branded demand was isolated from acquisition efforts

Rather than forcing growth, the account was allowed to scale naturally once efficiency was proven.

Final Takeaway

This case study shows how meaningful performance gains often come from fixing fundamentals rather than increasing spend or complexity.

By improving conversion quality, tightening Shopping structure, and removing internal inefficiencies, the account achieved stronger ROAS, higher revenue, and more consistent ecommerce growth. For niche ecommerce brands, this kind of disciplined optimisation is often the difference between flat performance and scalable results.

Contact Us