Prior experience, not Traffiy client work
I delivered this as part of my previous employment at a European digital agency. It was that agency’s client under its own contract, not a Traffiy engagement.
Eyewear · United States · 2022 · Senior Paid Marketing Strategist
Days from the season that makes their year, with nothing underneath the automation.
A US luxury eyewear label.
600%
ROAS over two peak months
-23.4%
Cost per click MoM
+30%
Clicks YoY
My role
Senior Paid Marketing Strategist
2022, at a European digital agency
The situation
Handmade frames, sold online and through the company’s own stores across several US states. Sunglasses means summer, and summer was days away when they handed over the Google Ads account.
The constraint
Everything was riding on automation with nothing underneath it to aim. That is survivable in a quiet month. It is not survivable in the one stretch of the year when every competitor in the category bids at once and the price of a click climbs accordingly. I had days to build and test, and anything I got wrong would be paid for at the highest prices the account would see all year.
What I did
- Gave the automation something to aim at. I built the search side out query by query, and arranged the product side around what the catalogue could realistically defend once prices climbed.
- Wrote the exclusion lists before launch rather than after. In a season this expensive, letting a system discover irrelevant traffic on its own is a bill you pay twice.
- Put the budget behind the queries that carry a buying decision, and behind the language people use when they are shopping at the top of the market rather than browsing a category.
- Stopped the account paying stranger prices for people who already knew the name. Returning visitors got their own path, and the brand terms got held at a sensible level instead of contested as though they were open ground.
- Read which slices were genuinely carrying the account, then narrowed what the automation was allowed to chase so the money went after demand we did not already have.
The outcome
Two months of peak season returned 600 percent ROAS. The account got cheaper to run while it grew, with cost per click down 23.4 percent month on month and return up 21.9 percent across the same period. Measured against the previous year it took 30 percent more clicks, converted 10 percent more of them, and brought in 17.2 percent more revenue.
What I took from it
Automation is not a strategy on its own. It optimises inside the structure you hand it, and where there is no structure it will spend contentedly into your own brand traffic. Most of the gain here came from building the thing the automation was supposed to be steering.
All work on this page was delivered during previous employment at another agency, as part of that agency’s teams and under its client relationships. It is not Traffiy client work. Brands are described rather than named, and every figure comes from an article published at the time.
Every figure above is quoted from the article published about this project at the time, and is recorded against its archived source. Happy to share the reference on request.