Data: July–October 2026 · 421 video ads from 21 brands · $10.3M in video ad spend
Part of our Meta Ads Statistics 2026 report.
The Short Answer
Across 421 Meta video ads that each spent at least $1,000 from July 6 to October 4, 2026:
- The median video was 29 seconds long. The middle half ran 18 to 42 seconds. Weighted by spend, the median was 34 seconds.
- Brands put most of their video spend into 16 to 45 second ads. About 61% of a typical brand's video spend went there.
- Length didn't predict which ads Meta spent the most on. Each brand's top 5 videos had a median length of 29 seconds. So did everything else.
- No length won on cost. Cost per purchase or lead stayed within about 6% of each brand's median in every length bucket.
- People watched about 3 to 5 seconds, no matter how long the ad was.
So the answer to "how long should a Meta ad be" is: as long as the message needs. The length matters a lot less than the first 3 seconds.
The Data
We pulled July 6 to October 4, 2026 data from the ad accounts we have visibility into, keeping brands that spent $50K or more over the period. For each account we looked at its top 50 ads by spend and kept the video ads that spent at least $1,000.
Brands often relaunch the same video under new ad IDs, so we counted each video once per brand and combined its results. That left 421 unique videos from 21 brands, with $10.3M in combined spend.
The brands are a mix of consumer products (food, supplements, pet, beauty, home) and apps and services (software, streaming, telecom, health). Brand names are anonymized.
Big accounts can drown out small ones, so most comparisons below are made within each brand first. For example, "1.03x" means an ad's cost per result was 3% higher than that brand's median video.
How Long Meta Video Ads Actually Are
| Length | Share of videos | Typical brand's share of video spend |
|---|---|---|
| 10 seconds or less | 10% | 12% |
| 11–15 seconds | 10% | 10% |
| 16–30 seconds | 36% | 36% |
| 31–45 seconds | 23% | 25% |
| 46–60 seconds | 12% | 9% |
| Over 60 seconds | 9% | 8% |
The 16 to 45 second range is where most of the money goes. About 1 in 5 videos was 15 seconds or shorter, and fewer than 1 in 10 ran over a minute.
Brands differed a lot. The median video length by brand ranged from 15 seconds to 80 seconds. A prescription health brand ran mostly long, explainer-style videos. A household brand ran mostly videos under 20 seconds. Both were spending real money on what worked for them.
Does Length Affect Performance?
| Length | Hook rate | Avg. seconds watched | Impressions that reach the end | CTR vs. brand median | Cost per result vs. brand median |
|---|---|---|---|---|---|
| 10 seconds or less | 24% | 3 | 9% | 0.91x | 0.95x |
| 11–15 seconds | 17% | 3 | 3% | 0.89x | 1.03x |
| 16–30 seconds | 21% | 3 | 2% | 0.98x | 1.03x |
| 31–45 seconds | 27% | 4 | 1.7% | 1.02x | 0.94x |
| 46–60 seconds | 27% | 5 | 1.6% | 1.15x | 1.00x |
| Over 60 seconds | 16% | 3 | 0.2% | 1.08x | 0.95x |
A few things stand out.
Cost per result was flat. Every bucket landed between 0.94x and 1.03x of the brand median. Short ads weren't cheaper. Long ads weren't either.
Spend didn't follow length. Within each brand, we ranked videos by length and by spend. The correlation was close to zero (a median of -0.05 across 21 brands). Each brand's single biggest video ranged from 5 seconds to 81 seconds, and the median was 28 seconds.
Almost nobody watches to the end. The median viewer watched about 4 seconds, and that barely changed with length. A 30 second ad gets about 15% of its length watched on average. A 60+ second ad gets about 3%.
Longer ads had slightly higher click-through rates. Ads in the 46 to 60 second range had a 1.15x CTR vs. their brand's median, but also a 1.21x CPM, so the cost per result came out even. More time to make the case seems to earn more clicks from the people who stay, but Meta charges more to show them.
Why Short Ads Don't Automatically Win
The common advice is "keep it under 15 seconds." The data doesn't back that up, for two reasons.
Most people leave at the same point anyway. If the typical viewer is gone by second 4, a 15 second ad and a 45 second ad lose the same people at the same moment. What's different is what the people who stay get to see. A longer ad has room for a demo, proof and an offer. A shorter one has to pick.
Meta optimizes for results, not views. It shows each ad to the people most likely to buy or sign up. A longer video that does a better job convincing the few who watch can win, even with lower completion.
There's one place where short ads look better on paper: hold rate. Meta counts a ThruPlay at 15 seconds, or at the end of the video if it's shorter. So an 8 second ad that gets watched to the end counts as a ThruPlay, while a 40 second ad needs 15 full seconds. That's why videos of 10 seconds or less showed a much higher hold rate (8.5% vs. about 3% to 6% for longer videos). It's a measurement quirk, not a sign they're working better. More on hook and hold rate in our hook rate benchmarks.
What This Means for Your Ads
- Pick length based on the message. A simple product with one clear benefit might need 10 seconds. A product that needs a demo, proof or an explanation might need 45. Both can win.
- Put the most work into the first 3 seconds. That's the part almost everyone sees. Test several openings on the same body before you test lengths.
- Front-load the important parts. Since most people leave early, the product, the main benefit and the offer should show up in the first 5 to 10 seconds, not the last.
- Test cutdowns, but don't expect miracles. A 15 second cut of a 40 second winner is worth testing, especially for Stories and Reels. Just don't assume it will beat the original.
- Run a mix of lengths. Meta rewards creative diversity, and different lengths reach different people. 17 of the 21 brands in this data spent real money on videos in at least four of the six length ranges.
- Judge on cost per result. Completion rate and hold rate will always favor shorter videos. Cost per purchase or lead is what decides a winner.
How We Use This at Biddyco
We don't make ads to a set length. Each concept gets the length its message needs. Then we put most of the effort into the first few seconds, and test several hooks on the same body to find the opening that works.
Across 100+ accounts spending $20M+ a month on Meta, that's what we've seen hold up. If you want help making UGC, motion or hybrid video ads that get the first 3 seconds right, book an intro call.
For more data like this, see how static and video ads compare, how long winning ads last and our full Meta Ads Statistics hub.
FAQ
How long should a Meta video ad be?
There's no single best length. In our data, the median video ad was 29 seconds, and most spend went to ads between 16 and 45 seconds. Cost per result was about the same at every length, so make the ad as long as the message needs.
Do shorter video ads perform better on Facebook and Instagram?
Not in our data. Ads of 15 seconds or less didn't have a lower cost per result than longer ones, and length didn't predict which ads got the most spend. Short ads do show higher hold and completion rates, but that's partly how Meta counts ThruPlays.
How long do people watch Meta video ads?
About 3 to 5 seconds on average, at every length. The median across all 421 videos was about 4 seconds. That's why the opening matters more than the total length.
What percentage of people watch a Meta video ad to the end?
Very few. About 2% of impressions reached the end of a 16 to 30 second ad, and about 0.2% for ads over a minute. Even ads of 10 seconds or less only reached the end on about 9% of impressions.
Should I make 15-second cutdowns of my video ads?
It's worth testing, especially for Stories and Reels. But in our data, 11 to 15 second ads didn't beat longer ads on cost per result. A cutdown works best when the original's key message already lands in the first 15 seconds.
Are long video ads worth it on Meta?
They can be. Ads over 60 seconds were a small share of spend (about 8% for a typical brand), but they performed in line with shorter ads on cost per result. They tend to work best for products that need explaining, like health products.
