CTR Decay Is Normal, but the Curve Shape Is Not

CTR almost always falls after the first burst of impressions, and that is expected. The speed and shape of the fall tell you more than the starting number.

A video that opens at 9 percent and settles at 4 over a week is behaving normally. The high early number came from loyal viewers, and the settle is the audience arriving. What matters is where and how it flattens.

Reading three curve shapes

A gentle slope that flattens high means the video kept finding new audiences without losing them. This is the shape you want, and it usually means the package and the content agree.

A steep cliff in the first day, settling very low, means the algorithm tested the video and the broad audience rejected the package. The early fans inflated the start; the strangers did not follow.

A slow, steady decline over weeks is simply decay as the topic ages and the algorithm moves on. That is not a packaging problem. It is a shelf life.

The first numbers are a greeting. The final numbers are the verdict.

What to log each week

• Day one CTR, day seven CTR, and day thirty CTR. Three points describe the curve.
• The gap between day one and day seven. A large drop means the loyal-viewer effect is strong.
• Where the number stops moving. That floor is your true packaging performance.
• Retention at the same checkpoints. A falling CTR with steady retention is a smaller problem than either alone suggests.

When decay is actually good

A video that decays slowly can out-earn a video that spikes. A steady 3 percent that holds for months beats a 7 percent that dies in a week once you count total watch time.

I stopped chasing the highest opening CTR after I noticed my best long-term videos never had the biggest day-one numbers. They had the flattest curves, and the curves compounded while the spikes did not.

Read the whole curve before you judge a package. The opening is a rumor, and the floor is the fact.