LinkedIn Video Marketing: Formats, Length, and What Gets Watched

Video is not a universal upgrade over text on LinkedIn. It is substantially better at three jobs (demonstrating something, conveying that a person is credible, and walking through a process) and worse than a well-written post at nearly everything else, because a post can be read in twelve seconds and a video cannot.

This is for B2B teams deciding whether video is worth the production overhead.

The Three Jobs Video Wins

Showing something happening. A workflow, an interface, a before and after. Describing a screen in prose is strictly worse than showing the screen.

Establishing that a person is worth listening to. Forty seconds of someone speaking conveys competence and credibility in a way no amount of writing does, which is why founder video works for early-stage companies with no other credentials.

Walking through a process where sequence matters. "Here is how I read a Schedule A" is a video. "Schedule A contains the broker name and commission" is a sentence.

Outside those, a written post usually wins on reach, on effort, and on the reader's willingness to engage at all.

Length

Thirty to ninety seconds for feed video. Under thirty rarely says anything; over about two minutes, completion drops sharply and the feed responds to that.

The exception is a genuinely useful walkthrough, where people who care will watch four minutes. But they are watching because they specifically want that thing, not because they were scrolling, so treat long video as a resource you link to rather than as feed content.

Upload natively. A YouTube link gets a fraction of the distribution of the same video uploaded directly, for the obvious reason that LinkedIn would rather keep people on LinkedIn.

The First Three Seconds, and Captions

Video autoplays silently in the feed. Most people never turn sound on.

So two things are non-negotiable. Burned-in captions, or at minimum LinkedIn's auto-captions checked for errors; an uncaptioned B2B video is being watched by almost nobody. And visible motion or text in the first frame, because a static talking head with no sound is indistinguishable from a paused video.

Put the claim on screen as text in the first two seconds. That is the equivalent of the first two lines of a written post, and it does the same job: it decides whether the rest gets watched.

Production Value Is Mostly Irrelevant

Phone footage with good audio outperforms studio footage with bad audio, consistently and by a wide margin.

Audio is the one thing worth spending on. A cheap lapel microphone does more than any camera upgrade. Beyond that, a clean background and reasonable light is enough, and over-produced B2B video often reads as advertising, which raises the viewer's defences.

The practical implication: do not let production overhead be the reason you do not ship. A 45-second phone video recorded once is worth more than a polished one that never gets made.

Video Post Copy Still Matters

A common mistake is treating the video as the whole post and writing "New video! 👇" above it.

The text is what appears in the feed and what gets read by people who will not watch. Write it as a standalone post that makes the point, with the video as the demonstration. Many people will read the text, not watch, and still take the point, which is a good outcome, not a failure.

Measure Completion, Not Views

A view on LinkedIn is counted after a very short watch time, so view count largely measures how many feeds the video appeared in.

Completion rate and comments are the real signals. A video with 600 views and 45% completion is doing better than one with 4,000 views and 8%, and only one of those is telling you the content landed.

And as with everything else on LinkedIn: the number that actually matters is inbound conversations from the right companies, not any engagement metric.