From 13.88 to 19 ROAS in a single month
A previous agency spent six months trying to push OnlineTopGarden past 19 ROAS and failed. Using disciplined A/B testing, we got there in one — without disrupting a profitable live campaign.
19.13×
ROAS achieved
1 mo
To hit the target
957
Conversions / month
+15%
Purchase value
The client
OnlineTopGarden is an e-commerce store with an extensive range of growing supplies — everything an indoor or outdoor grower needs in one place, so customers get the best results in the easiest way possible.
The challenge
This was one of the toughest briefs we've taken on. The account was already profitable at ~13 ROAS, set up by a previous provider — and the client wanted it pushed to 19–20 ROAS, a level the earlier agency had chased for six months without success. Getting 19–20 ROAS out of a mature ad account is genuinely hard, and the existing campaign couldn't be put at risk while we did it.
Our approach
Because the live campaign was already making money, the priority was to improve it without breaking what worked. So instead of editing the primary campaign directly, we ran a controlled experiment alongside it.
- Built a dedicated A/B testing campaign so changes could be validated against the live baseline with zero risk to existing performance.
- Iterated on the variants — testing structure, targeting and creative to find what actually moved ROAS.
- Rolled the winners forward once the test data showed a clear, sustained lift within the first month.
The results
What another agency couldn't do in six months, the A/B approach delivered in one. ROAS climbed from 13.88 to 19.13 — and spend actually came down while conversions and revenue went up.
| Metric | Before | After |
|---|---|---|
| ROAS | 13.88 | 19.13 |
| Ad spend / month | €5.07K | €4.24K |
| Conversions / month | 882 | 957 |
| Purchase value / month | €70.3K | €81.1K |
"At the start OnlineTopGarden was getting 13 ROAS — after we started the work we got 19 ROAS in the next month alone."