Ideal B2B video length: what the data shows

Wistia defines engagement rate as the average percentage of a video that viewers watch, which makes it a retention metric. For its 2026 report, the company analyzed more than 13 million videos hosted on its platform¹. Videos under 1 minute averaged 52%, and the number barely moves by video type: 46% for customer testimonials and 50% for product videos¹.

These are company videos placed on websites, product pages and galleries, not in a social feed. That brings the data closer to the B2B buyer, but the base mixes companies from every industry and has no country breakdown: it is global data.

The previous report, from 2025, shows the curve falling as length grows. According to Chief Marketer's coverage, viewers of videos under 5 minutes watched 47% of the video on average; 5 to 30 minutes, 38%; over 30 minutes, 21%³. In the 2026 data, Wistia also reports an 11% drop in engagement once a video passes the half-hour mark¹.

Share watched is not the whole story. Wistia itself notes that a video 10 minutes long with a lower engagement rate still racks up more total watch time than a video 1 minute long with a higher one⁴. On YouTube, which gives more reach to videos with more watch time, that difference matters.

Short video wins on retention and loses on action. In Wistia's 2025 data, 17% of viewers of videos longer than half an hour took the suggested action inside the video, versus 5% for videos 3 to 5 minutes long and 2% for shorter ones³. In the 2026 data, videos between 5 and 30 minutes averaged a 9% click-through rate¹.

The likeliest explanation is the intent of whoever presses play. A short video is often the first contact with a brand; a long video is sought out by someone already evaluating. Wistia CEO Chris Savage describes this split of attention in his interview with Chief Marketer: with an unknown brand, viewers invest little time; with a brand they trust, they invest a lot.

LinkedIn reached a similar reading with its own platform data. At the awareness stage, short videos delivered more than a 300% increase in completion rate; at the consideration stage, videos of 60 seconds or more drove more than 20% greater engagement⁵. The numbers are LinkedIn's, with no methodology published in the post, and they apply to that platform only.

Vidyard draws the same line for sales video: for first outreach to a prospect, the advice is to stay under 60 seconds; at the demo and proposal stage, the video can run long. In its data, only 20% of viewers finish a video longer than 20 minutes².

Format depends on where the video will play. On LinkedIn, vertical videos in 9:16 showed a 24% lift in click-through rate, and landscape 16:9 drove a 140% increase in completion rate, in the context of large-scale campaigns and connected TV⁵. This is single-platform data, released by LinkedIn itself.

LinkedIn has become the main channel for B2B video. In Wistia's 2026 report, 81% of teams post videos on LinkedIn and 76% on YouTube⁴. The same research shows 76% of teams adjust the aspect ratio by platform, and uploads of vertical HD video grew 24% over the previous year⁴.

LinkedIn video ads can run from 3 seconds to 30 minutes and autoplay in the feed muted by default, which is why the platform's page recommends captions⁶. In practice, format drives the cut: vertical is for the mobile feed, and horizontal stays on the website, YouTube and the big screen.

Our recommendation: don't pick a single format. Shoot with both framings in mind and produce a short vertical version for the feed and the full horizontal version for the website and YouTube.

The audience retention curve shows, second by second, how much of the audience is still watching. In YouTube Analytics, the audience retention report includes the intro, the percentage of the audience still watching after the first 30 seconds, and lets you compare the video against the typical retention of your 10 latest videos of similar length⁷.

The shape of the line tells you what happened. According to YouTube Help, a flat line means viewers watched that whole stretch; a gradual decline, interest fading over time; a spike, a part rewatched or shared; a sharp dip, viewers abandoning or skipping at that point. Key moments are highlighted only on videos with at least 60 seconds and 100 views⁷.

Wistia suggests splitting the curve into three parts: the nose, the first 2% of the video; the body, the middle 96%; and the tail, the final 2%⁸. Across the platform, the average loss in the nose is 4.9% for videos 1 to 2 minutes long and rises to 17.3% for videos 5 to 10 minutes long⁸. The study dates from 2020: use it as a rough reference, not a target.

Each part calls for a different fix. A drop in the nose is a weak hook: a long intro sequence, a slow setup, a promise that doesn't match the thumbnail. A drop in the body is a script problem: a repetitive or slow stretch. A drop in the tail matters when the call to action sits at the end, after wrap-up signals like music ending or a logo on screen.

The method we recommend uses the retention curve of the videos you have already published. The idea is to find the point where the audience leaves, compare against videos of similar length, and cut or reorder the script before the next one. Repeat the process for each platform, because a YouTube curve does not hold for LinkedIn or for the player on your website.

1. 01
   Split your videos by platform and by goal: awareness, consideration or decision. A good length for one goal is no yardstick for another.
2. 02
   In YouTube Studio, open the video's Analytics, Engagement tab, and look at audience retention. For your website player, open the retention heatmap in your hosting platform.
3. 03
   Write down the intro: how much of the audience was still there after the first 30 seconds⁷. Compare it with the typical retention of your 10 latest videos of similar length⁷.
4. 04
   Mark the point where the curve falls below half the audience. Everything after it is seen by a minority; the main message needs to come before.
5. 05
   Look for key moments and spikes. If the part viewers like most sits at the end, move it up front, as YouTube Help itself recommends.
6. 06
   Cut two versions of the same footage: a short, vertical, captioned one for the feed that ends at the drop-off point, and the full horizontal one for the website and YouTube.
7. 07
   Publish, wait for the new curve to build up, and compare the intro and the drop-off point with the previous version.

One caveat: a video with few views produces an unstable curve. With few people watching, each drop-off weighs heavily, and the conclusion becomes a hypothesis.

The next step fits in one editing meeting. Pick the published B2B video with the best retention, open its curve and find the point where the audience falls below half. That is the limit of the short version: everything essential has to fit before it, with the hook landing inside the first 30 seconds YouTube measures as the intro⁷.

Agree with your team on four production decisions for the next video: length, taken from the curve and not from gut feeling; the hook, delivering on the thumbnail's promise right away; format, vertical for the feed and horizontal for the website; and captions, mandatory on any version that will play without sound.