Video ad length: the runtime is a result, not a decision

Under fifteen seconds carries one claim and nothing else. It works when the claim is strong and the product is simple.

Fifteen to thirty seconds carries a claim and one piece of proof. That is the shape most direct response ads want.

Thirty to forty-five seconds carries a claim, proof and an objection handled. It is the longest most feeds tolerate from a brand they do not know.

Forty-five to sixty seconds needs a genuine story or a demonstration with steps. It also needs the first four seconds to earn the rest.

Beyond sixty seconds in a vertical feed you are asking for attention that has to be repaid. Most ads cannot repay it.

Pick by what your script contains rather than by a number somebody published.

Film up to three minutes of talking. That ceiling is deliberate. It is long enough to say everything twice and short enough that nobody rambles for an hour.

One batch of 100 credits returns a transcript-driven cut. The editing happens on the words rather than on a timeline.

Delete a line you do not want and the cut rebuilds around it. That is the mechanism that makes cutting to the claim quick.

Change the pace and the whole thing re-cuts. The ad gets shorter or longer without you choosing a target duration.

Then read the runtime off the result. If it came out at twenty-two seconds, the ad is twenty-two seconds.

Deciding on thirty in advance and padding to reach it is how ads get slower without getting better.

The pace lever does the work a duration target was pretending to do, and it does it on the whole ad at once.

Length arguments are usually an opening problem wearing a disguise. A strong opening buys more time than any duration rule grants.

Put a number, a name or a result in the first four seconds. Abstract openings do not survive being met at a scroll.

Write the opening headline yourself instead of accepting the drafted one. It is the highest yielding text in the ad.

Emphasise one word so it pops in the captions. That is the only tone of voice available with the sound off.

If the opening is weak, cutting from forty seconds to twenty does not fix it. It shortens the disappointment.

Fix the first four seconds, then stop worrying about the last twenty.

Export costs 20 credits per output minute, so length is nearly free on the export side.

Twenty seconds costs about 7 credits to export. Sixty seconds costs 20. The batch is 100 credits either way.

So a forty second ad is about 113 credits and a sixty second ad is 120. That is not a reason to choose one over the other.

Rendered presenters are the exception. At 320 credits a minute, going from forty to ninety seconds adds roughly 267 credits to a single ad.

Generated music is flat at 20 credits regardless of length, in 15, 30, 60, 90 or 120 second options.

So on a filmed ad, pick the length the script produces. On a rendered one, the credit meter is a real argument for brevity.

Every export is 9:16. A horizontal placement with its own duration conventions is a separate job.

The source take runs up to three minutes. A long demonstration gets filmed in parts and batched separately.

There is no timeline, so trimming a single frame is not on the menu. You trim the opening or the ending, and you delete lines.

Platform maximums move. Check the current limit for wherever the ad is going before assuming sixty seconds is available.

And a considered purchase sometimes needs more time than a feed will give it. That is a channel decision, with the landing page doing the explaining.

One habit is worth keeping regardless of length. Watch the finished export on a phone at arm width with the sound off.

An ad that feels short on a monitor often feels long on a phone. That gap is where most length arguments actually come from.

Watching muted removes the other illusion, which is that a good voice-over is carrying a section it is only decorating.

Then change the pace and re-cut. Nowhere else can you test three runtimes of the same ad in an afternoon for 20 credits each.