Beauty & Personal Care · Performance Marketing
Emberline Ethnics used programmatic to make retargeting worth paying for again
2.9x incremental ROAS, proven by holdout
A fragrance brand suspected its retargeting was buying customers it already had. A geo holdout test settled the argument and rebuilt the plan.
2.9x
Incremental ROAS
31%
Of retargeting revenue was incremental
17%
Revenue growth on flat spend
22
Matched test markets
The challenge
Emberline Ethnics spent 34% of its media budget on retargeting reporting a 9.2x ROAS. Leadership suspected the number was fiction but had no way to prove it, and the incumbent agency had no interest in running the test.
Strategy
Run a proper incrementality test before spending another quarter arguing about attribution, then rebuild the programmatic plan around audiences the brand does not already own.
Execution
We ran a 6-week geo-based holdout across 22 matched markets, suppressing all retargeting in the test group. We then rebuilt programmatic around prospecting: contextual placements on beauty editorial, CTV against a lookalike seed of high-LTV customers, and a tightened retargeting window of 7 days instead of 45.
Results
The holdout showed only 31% of retargeting-attributed revenue was incremental. We cut retargeting budget by 62% and moved it into prospecting, which returned a measured 2.9x incremental ROAS. Total revenue grew 17% on flat spend.
“They talked us out of spending money with them. That is when we knew we had the right partner.”
Farhan Qureshi — General Manager, Emberline
