Video Ad Length by Platform: What the Benchmarks Actually Say

There is a tidy rule that gets repeated in every paid-social thread: shorter ads are always cheaper, so cut everything to the bone. It sounds right because on most feeds it is right. But it falls apart on the one platform where the longest ad in your account can also be the cheapest per result. That paradox is the most useful thing to understand about video ad length, and it is the reason a single number can never answer the question for you.

Here is the short version of why. Length is not really a creative choice. It is a lever on the two numbers you pay for: CPM, which the auction sets partly on how completely people watch, and CTR, which depends on whether the viewer reached your ask before they left. The right video ad length is the one that protects both on a specific placement, and that number is different on a sound-off vertical feed than on a chosen, sound-on YouTube view. Let us walk through why, then go platform by platform with the benchmarks that back each range.

Why length moves CPM and CTR in the first place

Auction platforms sell retained attention, so they reward creative that keeps people watching. The signal they read most directly is completion rate and average watch time. A 12-second ad that 60 percent of viewers finish reads as healthier than a 45-second ad that 15 percent finish, even though the longer one delivered more total seconds of message. Completion is the currency, not duration.

That feeds straight into cost. Stronger completion tends to earn cheaper distribution, while weak completion gets taxed with a higher effective CPM. So length shapes CPM indirectly, through the completion rate it produces, rather than through some flat price per second.

CTR is the other half of the equation. A click can only happen once the viewer hears the ask, which is why the opening seconds have to stop the scroll before length matters at all. On a short ad, the CTA lands while most of the audience is still present. On a long ad, by the time you reach the CTA most viewers have already gone, so your clicks get divided by a much larger impression count. Same clicks, weaker CTR, because the denominator now includes everyone who left at second eight. The lesson is that length has no single optimum. It has an optimum per placement, decided by how that placement is watched.

TikTok ad length: 9 to 21 seconds, with the hook in second one

TikTok ad length lives at the short end of the scale because the format is sound-off by default, thumb-driven, and infinite. The viewer's baseline action is to leave, and they decide in roughly the first second whether to grant you more. The paid sweet spot for most accounts lands in the 9 to 21 second range. Below about eight seconds you often cannot fit a hook, a demonstration, and a CTA without it feeling clipped. Above about 25 seconds, completion tends to drop sharply and CPM climbs to match.

TikTok's own guidance supports the lower bound rather than the upper one. In its creative best-practice research, TikTok for Business has reported that ads of 21 to 34 seconds drove the strongest lift in its studies, while separately noting that a large share of top-performing ads run under 15 seconds. The practical read: aim for 9 to 15 seconds as the safe core, and only stretch toward the low twenties when a demonstration genuinely needs the extra beats. The feed punishes slow openers harder here than anywhere else, which is why what actually converts on TikTok hinges on the first frame.

Instagram Reels ad length and the rest of short-form vertical

Instagram Reels ad length behaves much like TikTok, with a small amount of extra patience. A range of 9 to 18 seconds works well, and Meta's delivery tolerates a couple of additional seconds more gracefully than TikTok's does. The Reels format rules that drive conversions still reward a clean, captioned cut that front-loads the payoff.

YouTube Shorts ad length sits a notch higher, around 15 to 25 seconds, because Shorts viewers tend to arrive a fraction more willing to watch. The same scroll physics apply, so the discipline does not change, only the ceiling. You can see the placement-specific traps in this guide to what works and what to skip on Shorts.

The common mistake across all three is treating these as short video. They are not abbreviated versions of a long ad. They are complete ads that happen to be brief: hook in second one, mechanism in the middle, one CTA at the end. That discipline extends to the frame itself, since producing a 9:16 vertical ad carries its own framing and pacing constraints. When you must cut a 30-second concept down to fit, remove the setup rather than compressing every section evenly.

Meta feed and Stories run on two different clocks

Meta is not a single placement, and its placements do not want the same length. The in-feed environment on Facebook and Instagram is scrollable but slightly more patient than a vertical short. Ads in the 15 to 30 second band tend to perform, with 15 seconds a strong default. You have room for a longer setup than on TikTok, but the trade still holds: the longer you go, the more completion you give up, and past about 30 seconds you are usually paying for seconds most viewers never reach.

Meta's published benchmarks reinforce the short side. Its advertising guidance recommends keeping in-feed video to 15 seconds or less to hold attention through to the message, and Meta has reported that the majority of mobile video viewing happens with sound off, which is why a captioned cut matters as much as the runtime.

Stories run on a harsher clock. They are full-screen and tap-to-skip, and the skip is immediate. Treat Stories like a vertical short at 10 to 15 seconds, hook first, no slow build. A 30-second Story ad mostly buys impressions for content nobody reaches. Because Meta auto-distributes across placements, a single mid-length cut of 15 to 20 seconds, captioned because sound is off by default, is a reasonable lowest-effort default. But when a placement spends real money, cut to its clock instead of letting one length serve all of them.

YouTube ad length: the platform where longer can be cheaper

YouTube is the exception that breaks the shorter-is-cheaper rule, and it is worth understanding precisely because it inverts everyone's instinct. The platform is lean-back, sound-on, and largely watched on bigger screens. The viewer chose to watch a video, and your ad is the toll. That single fact changes the economics, and the format you pick decides the length.

• Bumper ads run a hard six seconds. They are non-skippable with a fixed cap, so you get one idea and one frame of branding. They suit reach and reminder, not explanation.
• Skippable in-stream is the interesting case. The viewer can skip after five seconds, and on a cost-per-view setup you are typically charged only when someone watches past 30 seconds or interacts. Google's documentation confirms that for skippable in-stream you pay when a viewer watches 30 seconds, the full ad if it is shorter, or engages with it. That billing model rewards a longer ad: 30 seconds to a few minutes can work, because the people who keep watching self-select into your warm audience and the people who skip cost you nothing. A 15-second skippable ad leaves the format's main advantage unused.
• Non-skippable in-stream gives you a guaranteed 15 to 20 second window, so use the whole thing. About 15 seconds is the practical target.

The first five seconds still carry the full load on skippable inventory, because that is the free preview before the skip button appears. Unlike a vertical feed, though, length beyond that point is not automatically a tax. This is the paradox from the opening: your longest YouTube ad can post your lowest cost per result.

LinkedIn ad length: short, because the impressions are expensive

LinkedIn CPMs typically run well above consumer platforms, often several multiples higher, so every wasted second costs more in absolute terms. The feed is professional, scrollable, and frequently watched with sound off. People are there to skim, not to settle in. LinkedIn's own video guidance points toward brevity, and its data has shown that shorter videos under 30 seconds tend to report higher completion rates.

Keep it tight: 15 seconds is a strong target and 30 seconds the practical ceiling. B2B buying cycles are long, but a feed video rarely closes the deal. Its job is to earn one click or one moment of recognition, which is the core of any B2B LinkedIn video playbook. Front-load the relevance, name the audience plainly such as "for RevOps teams running quota", make the value legible without sound, and place the single CTA early enough that skimmers catch it. Because impressions here are pricey, overshooting on length costs you more than it would anywhere else on this list.

A simple rule for choosing video ad length

You can pick the right length on any placement by answering three questions in order. First, how is this placement watched? Sound-off and scrollable means short and hook-first, which covers vertical feeds, Stories, and LinkedIn. Sound-on and chosen, which means YouTube in-stream, lets you run longer. Second, does the format charge me for skipped seconds? On feeds you pay per impression regardless of watch time, so every second past the CTA is dead weight and you should cut tight. On YouTube cost-per-view you pay only on a real view, so length is allowed. Third, can the message survive being shorter? Default to the short end of the range and add seconds only when a demonstration genuinely fails without them. Length should be the minimum that lands the hook, the mechanism, and one CTA, with that CTA placed where most of your completion curve still is.

Translated into numbers, the targets look like this:

• TikTok: 9 to 15 seconds, vertical 9:16, hook in second one.
• Instagram Reels: 9 to 18 seconds, vertical 9:16.
• YouTube Shorts: 15 to 25 seconds, vertical 9:16.
• Meta feed: 15 to 30 seconds, 1:1 or 4:5, with 15 seconds as the default.
• Meta Stories: 10 to 15 seconds, vertical 9:16.
• YouTube skippable in-stream: 30 seconds to two minutes, 16:9, first five seconds do the work.
• YouTube bumper: exactly six seconds, 16:9, one idea.
• LinkedIn: 15 seconds target, 30 seconds maximum, 16:9 or 1:1.

The pattern running through the whole sheet is consistent. The more the viewer chose to be there, and the more they can hear, the longer you are allowed to run. The more they are scrolling with the sound off, the more aggressive the cut needs to be.

The catch is production, not strategy

Read the sheet again and a problem appears. The correct move is not one ad at the right length. It is the same concept cut to several lengths, one per placement: a 12-second TikTok cut, a 20-second Meta-feed cut, a 6-second bumper, a 45-second YouTube version. That is four edits of a single idea before you have tested even one hook variant. Most advertisers skip this. They ship one 30-second master everywhere because re-cutting by hand is slow, then conclude a platform "does not work" when in fact they fed it the wrong length. The creative was rarely the problem. The cost of getting length right per channel was, which is why iteration speed is the real moat in paid video.

Common questions about video ad length

What is the best length for a video ad?

There is no single best video ad length. It is set by the placement. For scrollable, sound-off feeds such as TikTok, Reels, Stories, and LinkedIn, aim for 9 to 21 seconds with the hook in the first second. For the Meta feed, 15 to 30 seconds. For YouTube skippable in-stream, longer is fine and sometimes better, anywhere from 30 seconds to a couple of minutes, because you mostly pay only for views people choose to keep watching.

Are shorter video ads always cheaper?

On feed placements where you pay per impression, shorter usually protects completion rate and therefore CPM. On YouTube cost-per-view inventory the opposite can be true, because you are charged only when someone watches past the threshold or interacts, so a longer ad served to a self-selecting audience can post a lower effective cost than a forced-short one. The billing model of the format decides, not the raw number of seconds.

Should I run the same video ad length on every platform?

You can use one master as a starting point, but it leaves return on the table. Each platform watches at a different length and aspect ratio, so a single cut will be too long for some placements and miscropped on others. The higher-return approach is to re-cut one concept per placement against its clock, then test variations from there.

Producing a TikTok-length vertical, a Meta-feed square, and a longer YouTube version from one idea is the per-placement work that stops most teams from getting length right. Aitachyon generates a captioned video ad from a single concept and exports it across 9:16, 16:9, and 1:1, so cutting a length per channel takes minutes rather than a fresh round of edits.