Creative Fatigue vs Audience Fatigue

I’ve read a lot about creative fatigue in paid media, but not so much about audience fatigue.

This is something I experienced recently with a client, and I feel that it warrants a larger conversation.

I’ve been working with a client in the health/fitness space for some time now, and as we’ve run multiple tests across different service lines, I started to notice something interesting.

When running ads against a cold audience, performance in terms of CTR, view rates, CPL, and CAC would start to drop off after a few weeks, or even a few months depending on the service line.

My initial thinking was that this was most likely ad fatigue. After all, ad performance was exhibiting all of the typical signs I’ve seen in the past when dealing with ad fatigue. As a result, we started testing new creative.

The only trouble was that we didn’t see an improvement in performance. We thought maybe these new concepts just didn’t hit home the way we wanted, so we tested some more ads.

The same thing happened.

Rinse and repeat, but regardless of the new creative we didn’t see performance return to the peak levels we had seen shortly after launching these campaigns.

When digging further into some customer insights, we found that the most recent leads we were generating claimed they had never heard of this client before. This made me think.

If we were experiencing ad fatigue, then in theory brand recall should’ve been somewhat decent, but this wasn’t the case.

Further analysis revealed that we most likely had exhausted all of the “warm” prospects within this cold audience.

When digging into self reported attribution (SRA) data, I found that many of the leads we had generated before the “ad fatigue” kicked in cited another marketing channel as the place where they heard of my client’s brand.

My hypothesis is that Meta had access to strong enough signals to target users who were most likely to be quality customers early on, but this user pool was finite. Basically, Meta found all of the “warm” users quickly, but we ran through this warm audience and that’s when performance started to decline.

Ad fatigue wasn’t the issue, we had just inadvertently been targeting a warm audience. The change in performance was more so the result of showing ads to a warmer audience first, then moving to a colder audience. 

No audience settings were changed at the ad set level during these tests.

After discovering this, we’ve pivoted our strategy to focus on more structured audience layers to reduce our reliance on Meta’s algorithm.

This was a great case study for me to better understand how algorithms behave, why audience structure is important, and to reaffirm previous analyses that showed me it takes several interactions with a brand before people are ready to raise their hand.

Have questions, considerations, or critiques? I’d love to hear them! If you’re reading this via email, just hit respond. Otherwise, you can find me on LinkedIn.