The short version
- ROAS rose from 451% to 700% (+55%) for April–August 2026 vs the same months of 2025. Conversion value grew 42% while ad spend fell 9%.
- The 500% ROAS target has been met 11 months in a row. Before the handover it was met in 1 of 4 full months.
- Main levers: moving budget from low-return EU tests to the UK, running low-bid Shopping next to Performance Max and scaling a newly launched product category.
The client and the challenge
A UK-based e-commerce brand selling kitchen and home appliances, with a product catalogue and Google Ads campaigns across the UK and four EU markets. I took over the account in August 2025, first together with a colleague and then on my own.
Client goals
ROAS of 500% or higher, cost per purchase under an agreed ceiling, and continued tests beyond the UK.
The challenge
ROAS hovered around 430–450%, below target, and a third of the budget went to EU markets that returned about half of what the UK did.
Starting point
Baseline: April–August 2025, the first period with purchase-only conversion tracking.
Below target. The 500% goal was reached in only one of the four full months before the handover.
Budget in the wrong place. A third of spend went to four EU markets at a blended 210%, against 565% in the UK.
Almost everything in Performance Max. PMax took ~87% of spend, and generic non-brand Search returned only about 260%.
A new product line was coming in November 2025 and needed traction without draining proven campaigns.
What was done, and why
Budget allocation came first because it was the fastest and safest lever. Structure and growth levers followed once the account was stable.
Spend shifted from weak EU campaigns to the UK, and generic non-brand Search was cut from ~5% to ~2% of spend. The EU markets stayed live as capped tests.
UK share of spend 68% → 90%Standard Shopping campaigns in all five markets at minimum bids of a few pence. At that level they only win cheap, low-competition auctions PMax leaves on the table: little cost, extra sales.
~3.5% of spend at ~970% ROASIt started in a limited test campaign so learning costs didn’t hit the core lines, then got its own UK Shopping-led PMax once the data supported it.
~11% of spend at 530%8 hero products turned into 51 listings with titles and descriptions aimed at different search intents, to widen coverage across the catalogue.
Dedicated UK PMax for the second product line, category-level tests in Germany and Spain, and a relaunch of the main UK campaigns as fresh copies. Bid strategies were chosen per campaign by testing.
Refreshed campaign: ~980% in its first five weeksBefore → after
Before = April–August 2025, after = April–August 2026. Shares are of Google Ads spend.
Year over year: April–August 2026 vs 2025
Same months a year apart, purchases only. Fewer impressions (−32%), but more relevant traffic: CTR +44% and click-to-purchase rate +27%.
ROAS against the 500% target
Conversion value ÷ spend. The vertical line marks the account handover. Q3 26* = July–August only.
Year-over-year change by metric
April–August 2026 vs April–August 2025
Conversion value vs spend, indexed
Monthly, April–August 2025 average = 100
Where the budget goes
Share of spend and ROAS by market, April–August
This was not just budget shifting: the UK itself improved from 565% to 731% ROAS (+29%).
July–August 2026: same budget, 60% more value
What didn’t go to plan
Summer heatwave and stock-outs
Promotions underperformed and top products sold out with long restock times. Sales dipped in June, but ROAS held at 660%.
Key takeaways for e-commerce brands
- Fix allocation before you optimize. Moving budget to the market that already paid back was the biggest and fastest lever.
- Low-bid Shopping works well next to PMax. It costs very little and picks up sales PMax leaves behind.
- Treat new markets and products as capped experiments. Own campaigns and a clear exit rule, so they don’t dilute what works.
Running Google Ads for an e-commerce brand?
If your ROAS has stalled or budget is spread across markets that don’t pay back, I’m happy to look at your account and share what I’d change first.
Free and with no obligation: with read-only access I review your account before the call and bring the main changes I would make first.