Where Should You Put the CTA in a Video? (And When It's Killing Your Conversions)

Most advice about video CTAs focuses on wording — what to say, what button color to use, what verb to lead with. Almost none of it addresses the more basic problem: a perfectly worded CTA does nothing if it's placed after the point where most of your audience has already left. Placement isn't a minor detail sitting underneath the "real" CTA question. For a lot of videos, it's the whole ballgame.

The default placement is often wrong by default

The standard pattern — say the pitch, then close with a CTA in the last five to ten seconds — comes from long-form formats where finishing the video was a reasonable assumption about viewer behavior. That assumption doesn't hold on most modern video platforms. If your retention curve shows that only a third of viewers are still around by the 80% mark, an end-card CTA is only ever reaching that surviving third — and the two-thirds who left never saw it at all, no matter how well it was written.

This is easy to miss because a CTA at the end feels structurally correct — it's where a pitch, an argument, or a story naturally concludes. But "where it feels narratively correct" and "where your actual audience still is" are two different questions, and only the retention curve answers the second one.

Matching CTA placement to your retention curve, not a template

The more reliable approach is to look at where your specific video's attention actually holds, and place the ask inside that window rather than after it.

If retention falls off a cliff early and never recovers, a single end-card CTA is close to worthless — the audience that would act on it is already gone. In this case, an earlier, lower-key CTA (a mention, an on-screen callout, a spoken line woven into the content itself) placed before the drop-off point reaches far more of the people who were ever going to see it.

If retention holds relatively flat through most of the video with a gradual decline, a CTA near the end still reaches a meaningful share of viewers, and can work as the primary ask — since a flat curve means the audience that's going to see the end is a fair proxy for the audience overall.

If there's a specific point where retention dips and then partially recovers (common after a slow section followed by a stronger one), placing a CTA right after the recovery point can catch an audience that's re-engaged and paying closer attention, rather than one that's already mentally checked out from the dip.

Multiple, lighter touches usually beat one big ask

Especially for longer videos, a single CTA moment asks a lot of one specific slice of your retention curve. Spreading the ask across two or three lighter mentions — an early one for viewers who won't make it to the end, a mid-video one, and a closing one for those who do — tends to outperform betting everything on one placement, precisely because it doesn't assume where your audience will be when the moment arrives.

Know your curve before you decide where to ask

None of this is guessable from watching your own cut — you already know the ask is coming, so you don't experience the drop-off the way a first-time viewer does. Axony generates a predicted, second-by-second attention and retention curve for your edit before you publish, so you can see exactly where your audience is statistically likely to still be watching — and place your CTA where it will actually be seen, instead of where convention says it should go.