The setup. Native widget campaign, nutra offer, broad US desktop. One creative carried it for 11 days. Figures are illustrative.
The move. When ROI slid from 1.4 to 0.9 I assumed creative fatigue and swapped in six fresh angles. They all underperformed the "tired" one.
The numbers. Original creative lifetime: $14,000 spend, 1.28 blended ROI. The replacements burned $1,900 at 0.71 before I pulled them. The original at the SAME frequency cap was still doing 1.1 — I'd been letting frequency climb to 7+ on a finite audience.
The lesson. On a capped traffic source, falling ROI isn't always the ad wearing out. It's the audience saturating. Same creative, same people, fifth time — of course response drops. I was treating a reach problem like a creative problem.
What I'd do differently. Separate the two signals. If a new creative to the same audience also tanks, it's saturation, not fatigue — widen geo or add placements, don't churn creatives. I now log frequency alongside ROI on every line so I can tell which wall I actually hit.
Arb Files
@ArbFiles
The creative didn't die. The frequency did.
This post was published in the Telegram channel Arb Files (@ArbFiles). All its posts are on the channel page.
Open on Telegram