How Long Should a Promo Video Be?

The short answer: as long as the job requires and not one second longer. A promo video that interrupts a feed has a different job than one that greets a visitor who already clicked your ad — so they should not be the same length. Here is the placement-by-placement guidance, the reasoning behind it, and the word budgets that keep each length honest.

The quick reference

Length follows attention, and attention follows intent

The single principle underneath every row of that table: match the video’s length to how much attention the viewer has already agreed to give you.

• Interruption contexts (ads, feeds): you have borrowed attention. The viewer did not ask for this. Fifteen to thirty seconds is the respectful maximum, and the first three seconds decide everything — front-load the reason to stay.
• Consideration contexts (your website, a docs page): the viewer clicked something to get here. Thirty to ninety seconds is welcome if every second earns its place.
• Committed contexts (a launch event, an email to fans, a sales call follow-up): minutes are fine. These viewers want depth, not brevity.

The word budget test

Narration runs at roughly 150 words per minute. Before debating seconds, count words in the script:

• 15 seconds ≈ 35 words — one idea, stated once
• 30 seconds ≈ 75 words — one idea, with one proof
• 60 seconds ≈ 150 words — one idea, proof, and a short tour
• 90 seconds ≈ 220 words — a full narrative arc

If your message will not fit the budget, the fix is almost never a longer video — it is a smaller message. Our script templates are built around those budgets, beat by beat.

A useful heuristic: one message per thirty seconds

Viewers reliably retain about one idea per half-minute of promo. A 30-second ad carries one message. A 60-second explainer carries one message plus how-it-works. A 3-minute launch film carries a story — which is still, structurally, one message wearing different outfits. When a draft feels long, it is usually not too many seconds; it is too many messages sharing them.

Cutting a 60 down to a 30 (without breaking it)

The most common real-world task is not writing a fresh short script — it is shrinking an existing one. Three moves preserve the argument while halving the runtime:

1. Cut setups, keep payoffs. Long cuts explain the problem and then show the fix; short cuts open on the fix and let the viewer infer the problem. The hook becomes the payoff line.
2. Collapse proof to its strongest single item. Three testimonials become one; a feature tour becomes the one feature people actually buy.
3. Never speed up the read. A rushed voiceover is the audible version of fine print. If the words don’t fit at a natural pace, remove words — not pauses.

Run the mute test on the short cut afterwards: with captions on and sound off, the 30-second version should still land its one message. If it only works with narration, the cut leaned on speed instead of selection.

When longer genuinely wins

Long videos are not a failure state. They win when the viewer’s question is big:

• Demos for considered purchases. Nobody buys week-long-implementation software off a 15-second clip. A 90-second guided tour that honestly shows the product beats a montage.
• Founder stories. Trust takes narrative time. These run 1–3 minutes and belong on pages and channels where people chose to be.
• Testimonial compilations. Three customers at 25 seconds each is 75 well-spent seconds.

Stop choosing — make the set

The practical resolution of the length debate is that you no longer have to pick one. Production used to be expensive enough that teams made a single 60-second video and forced it everywhere: too long for ads, too shallow for the demo. With AI production the marginal cut is cheap — on UltraVora you can render anywhere from 15 seconds to three minutes, in widescreen, vertical, or square (see which aspect ratio to use where), so the sane workflow is a master cut plus purpose-sized versions: a 60-second hero for the site, a 30-second ad, a 15-second teaser.

The honest checklist

• Where will this video actually run? Length follows placement.
• Does the script fit the word budget without rushing?
• Is there exactly one message per 30 seconds?
• Would you watch to the end if it were not your product?
• Could the same script ship as a shorter cut for borrowed-attention channels?

If you want to feel the difference rather than theorize about it, generate the same promo at 15, 30, and 60 seconds and watch them back to back — the full workflow guide shows how, and a free session is enough to run the experiment on your own product.