LinkedIn Video Format in 2026: Specs, Length and Performance

Most LinkedIn videos don’t underperform because of bad content — they underperform because the format is wrong and the strategy stops at publishing.

Getting LinkedIn video right in 2026 means more than picking a size and a length. It means uploading a file the platform actually accepts, choosing a shape that earns screen space on a phone, and making a handful of decisions — thumbnail, caption, first hour — that determine whether anyone presses play at all.

This guide covers all of it: the specs LinkedIn publishes, the aspect ratios worth using, how long a video should run, and the performance tactics that decide who watches. Where a figure appears here, the source it came from is named next to it. Where a widely repeated number turned out to have nothing behind it, it isn’t in this post.

LinkedIn’s Actual Video Specs

Technical specs aren’t interesting, but getting them wrong costs you the post. The figures below are LinkedIn’s own, from its published video specifications — not a rule of thumb.

The one change worth flagging: MOV and AVI used to work and don’t any more. That is the single most common reason an upload fails outright.

Aspect Ratio: Which Shape Earns the Most Screen

Different aspect ratios serve different purposes. The practical question isn’t which one LinkedIn prefers — it accepts all of them — but how much of a scrolling phone screen your video occupies, and whether the same file still reads on a desktop monitor.

Square and portrait: the default

Square (1080x1080) and portrait (1080x1350) take up more vertical space in a phone feed than landscape does, which is the whole argument for them. They suit talking-head content, quick tips and announcements, and they degrade gracefully on desktop.

Vertical: mobile-native, desktop-awkward

The 9:16 ratio gives you the full-screen mobile experience people expect from short-form video elsewhere. The trade-off is real: vertical video looks cramped and letterboxed on a desktop monitor. Use it when you know your audience is on a phone.

Landscape: when you need the width

Traditional 16:9 earns its place when horizontal space carries the meaning — screen recordings, charts, slides, interview two-shots. Text stays readable and demonstrations don’t feel cropped. It just wins less attention in a mobile scroll.

How Long Should a LinkedIn Video Be?

Almost every guide tells you to keep videos under 90 seconds. That advice isn’t wrong, but it misses the point. The number that matters is not your video’s length — it’s how much of it people actually watch.

A 90-second video where viewers drop off after 20 seconds performs worse than a three-minute video people finish. LinkedIn reports completion rate and average watch time in your post analytics, and those are the numbers to design around. Your goal isn’t the shortest possible video. It’s a video where length matches value.

A starting map, not a rule:

• 15–30 seconds: quick tips, announcements, teasers. One idea, no preamble.

• 30–90 seconds: product demos, case-study highlights, educational snippets. Enough room for a complete thought — problem, solution, next step.

• 2–5 minutes: tutorials, interviews, comprehensive breakdowns. Fewer people start; the ones who do are the ones you want.

• 5–10 minutes: deep dives and panel excerpts. Ten minutes is LinkedIn’s hard ceiling for feed video.

Read your retention curve instead

Most videos lose viewers in a predictable shape: a sharp drop in the first few seconds from people who never intended to watch, a gradual decline through the middle, and sometimes a small uptick at the end from people who committed and want closure.

You can see this in your own analytics. If you consistently lose most viewers around the 45-second mark, you have two options: cut the video there, or restructure so your most valuable content sits before the drop-off. Some creators front-load the best insight and use the second half for context and examples. Others build tension and pay it off at the end. Both work — but they need different structures, so find out which curve you actually have before choosing.

When longer earns its place

Three to five minutes can beat 60 seconds for deep dives into complex topics, case-study breakdowns and interview clips with recognised experts. Those formats attract people actively seeking depth rather than scrolling for quick hits. The trick is setting expectations: if your caption and thumbnail promise a comprehensive breakdown, viewers arrive ready to invest the time. Don’t bait people with a quick-tip hook and then deliver a lecture.

The Thumbnail Is the Headline

We obsess over scripts, b-roll and audio quality, then accept whatever frame the platform picked — usually mid-sentence, mouth half-open. That is backwards. Most videos are decided before anyone presses play: on the thumbnail, on the opening line of the caption, on whether the post looks different enough from the fifteen other things in the feed.

Treat your thumbnail as a headline, or a book cover. Its job is to open a gap between what someone can see and what they need to know. A graph with only half the data visible. Your face reacting to something off-screen. Text that poses a question without answering it. Don’t summarise the video — make scrolling past feel like missing something.

The first-frame test

Videos auto-play muted as people scroll. You can’t control when that happens, but you completely control what sits in the frozen first moment.

Scroll your own feed and notice which video thumbnails make you stop. They’re rarely polished or corporate. They’re specific. They show something happening, not someone about to talk about something that might happen eventually.

Your opening frame should work as a standalone image. If someone saw it with no context at all, would they want to know what it’s about? That is the whole test.

Thumbnail checklist:

• High-contrast colours that survive a fast scroll

• A human face, when it’s relevant to the content

• Readable text overlay carrying the benefit or the question — not the answer

• Something in motion or mid-action rather than a static portrait

• An honest representation of the video — a thumbnail that oversells kills completion rate

• Consistent visual branding across your thumbnails, so regulars recognise you

Assume the Sound Is Off

You are making videos for silent viewers whether you like it or not. LinkedIn’s own research found that only 37% of people watching B2B video on mobile typically listen to it, against 56% on laptop or desktop, and that 21% of mobile B2B video viewers have their sound permanently turned off. Widely quoted figures of 80% or 85% for silent viewing circulate without a traceable source; LinkedIn’s own numbers are lower, and they still make the point.

Your video has to communicate value without audio. That does not mean subtitling every word you say. It means visual storytelling: show the problem, show the process, show the result. Use text overlays for the key points, not as a transcript of your monologue. When people do turn the sound on, it should add depth to something they already understand. The core message lands in silence or it doesn’t land at all.

Captions are an engagement tool, not just accessibility

Captions matter for viewers who are deaf or hard of hearing — that alone makes them non-negotiable. They also serve everyone watching in a quiet office, on a train, in a waiting room. LinkedIn’s own video guidance recommends subtitles for exactly this reason, and tells creators to get straight to the point in the opening seconds rather than warming up.

LinkedIn’s auto-captioning has improved, but it still mangles industry jargon, brand names and technical terms. Uploading your own SRT file takes about ten extra minutes per video and removes the problem. Poor captions cost you credibility; absent ones cost you the audience.

Captions also make a video skimmable, which sounds like a bad thing and isn’t. You don’t want everyone watching — you want the right people watching. Captions let a qualified viewer find the moment that matters to them, which raises average watch time among the people who actually stay.

Text overlays versus full captions

Full transcription supports accessibility and skimmability. Selective text overlays highlight key phrases on screen and carry sound-off viewing. The strongest videos use both. Make overlays large enough to read on a phone, high contrast, and on screen long enough to finish reading — and don’t overlay every sentence. Highlight the insights, the numbers, the memorable phrases, and give the viewer’s eyes somewhere to rest.

The Caption Is the Pre-Hook

Most advice tells you to hook viewers in the first three seconds of the video. True, but incomplete — you need to hook them before they start watching at all. That job belongs to your caption.

The best video posts use a caption that opens a knowledge gap. It hints at the insight without revealing it. It poses a problem the video solves, or makes a claim the video proves.

What doesn’t work: “In this video, I share three tips for better presentations.” That’s a description, not a hook. It tells me what I’ll get but doesn’t make me need it. Compare: “I watched our sales team lose three deals in a row with the same presentation mistake. Here’s what we changed.” Now I want to know what the mistake is, and whether I’m making it.

The pattern interrupt

LinkedIn feeds are text-heavy. Video is already a pattern interrupt, and you can amplify it with layout. Start your caption with a single sentence — not a paragraph. Direct, specific, maybe a little provocative. Then a line break. Then the video. Then the rest of the caption underneath.

This makes the video the focal point. Readers see one line, the thumbnail, and make a split-second decision. Everything below is for people who already decided to engage.

Don’t let your caption cannibalise your video

Long captions that essentially transcribe the video eat their own view count. The logic sounds reasonable — some people prefer reading, so give them both — but you are asking your audience to choose between two pieces of content in the same post. Most will skim the caption, feel they got the gist, and scroll on.

The caption and the video should do different jobs. The caption gives context, poses the question, names the framework. The video gives the detail, the examples, your personality. They complement each other; they don’t duplicate. If someone could get full value from the caption alone, there is no reason to press play.

One pattern that works well: use the caption to ask the questions the video answers. Readers mentally check off the ones they struggle with, and if even one lands, they watch. That turns the caption into a relevance filter — only people who need the content engage with it.

Quick checklist before you post:

• Does the caption pose a question or problem the video solves?

• Does it avoid revealing the full answer?

• Is the first sentence standalone and curiosity-creating?

• Is there zero duplicate information between caption and video?

• Does the caption set context while the video delivers the payoff?

Upload Natively, and Put the Link in the First Comment

LinkedIn’s algorithm has one straightforward goal: keep people on LinkedIn. Post a YouTube link and you are asking it to promote content that pulls users off the platform. Native uploads auto-play, display larger on mobile, and give you the full analytics set — completion rate and average watch time among them, which you do not get for an embedded external player.

External links still have a job. They are conversion tools: a webinar registration, a product demo, a long-form interview hosted elsewhere. They are not engagement tools.

You can have both, just not in the same slot. Upload the video natively, then post the external link yourself as the first comment, immediately after publishing. The algorithm gets native content; anyone who wants depth, resources or a next step has a path to it, sitting near the top of the comments where engaged viewers are already looking.

The First Hour

LinkedIn weighs early engagement heavily when deciding how far to distribute a post, which makes the hour after you publish disproportionately important. Give close connections a heads-up. Answer every comment quickly and substantively — not “Thanks!” but a real reply that extends the conversation and shows later readers that someone is home.

On timing: there is no universal best time to post. B2B decision-makers in New York browse differently from founders in San Francisco or consultants in London. Start from your own analytics — when do your text posts get the most engagement? Post video into those windows. We’ve looked at the general patterns separately in our guide to the best time to post on LinkedIn, but treat any published window as a hypothesis to test against your own audience, not a rule.

The strategic first comment

Comments tell the algorithm your content started a conversation, but most posts don’t generate them on their own — people watch, react, and move on. Seed it yourself. Post the first comment within minutes of publishing, and make it specific.

“What do you think?” is lazy and it doesn’t work. Vague prompts create decision paralysis; specific ones create conversation:

• Educational video: “Quick question on [specific step from the video] — have you tried this, and what happened?”

• Story video: “Has anyone else dealt with [specific challenge]? What was your approach?”

• Opinion video: “I know [main point] isn’t universally accepted. What’s the strongest argument against it?”

You are giving people permission to respond in a specific way, which is most of what makes them respond at all.

What Our Own Data Says About Video Versus Text

There is a claim in wide circulation — and it used to be on this page — that text posts get two to three times more engagement than video on LinkedIn, and that the algorithm increasingly favours text for organic reach. We measured it, and it is the wrong way round.

In our LinkedIn engagement benchmarks — 100,459 posts from 1,054 connected professional accounts — the median engagement rate by format comes out as text-only 1.54%, video 2.78%, document 2.67%, image 2.94%. Video beats text-only by roughly 1.8×. The finding also survives a within-account control: comparing each person against their own posting history, video beat that same person’s own text-only posts for 79.3% of the 328 accounts we could test.

Two honest caveats. Image, video and document are statistically indistinguishable from each other in our data — their confidence intervals overlap — so the real finding is any media beats no media, by roughly three times, rather than video specifically winning a format race. And this is correlational: people who go to the trouble of making a video may simply be putting more thought into the post.

What that changes in practice: don’t repurpose a video into a text post because text performs better. It doesn’t. Repurpose because one recording can answer several different questions for several different audiences — and when you do turn a video into a written post, pull a frame or a chart out of it and attach that, because a post with media beats a bare wall of text.

Repurposing: Reframe, Don’t Repost

Reposting the same video weeks later feels spammy and costs you credibility. Smart repurposing changes the context, not the timing. You are not reposting the content — you are using the same footage to answer a different question.

Say you made a video about handling difficult client conversations. The original post framed it as a communication skill. Three weeks later, the same footage with a new caption becomes a client-retention piece. Different audiences need different context; the underlying content is unchanged.

Think about the angles before you even film. A single piece of footage about your morning routine can be framed as productivity advice, as a mental-health argument, or as time management. You are meeting your audience where they are rather than asking them to remember what you posted a month ago. Our guide to content repurposing goes further on the strategic version of this.

Another move: pull the strongest 30 seconds out of a longer video and post it as a standalone piece. That isn’t duplication — it’s unbundling. A three-minute breakdown of four delegation mistakes contains at least four shorter posts, each aimed at a different pain point.

One practical note: to repurpose your own videos you need the files. If you need to retrieve something you posted straight to LinkedIn, ContentIn’s LinkedIn Video Downloader lets you paste the post URL and download the file so you can clip, reframe and reuse it.

Turning a video into a written post

A repeatable structure, rather than a formula to follow slavishly:

1. Hook: the single most compelling line from the video

2. Context: one or two sentences of background

3. Key points: three to five takeaways, broken up so they’re scannable

4. Supporting detail: the example or number that made the point land on camera

5. Close: a specific question, not “thoughts?”

6. Media: a still frame, a chart or a clip — media posts outperform bare text in our own data

Which Video Metrics Actually Matter

Views alone tell you almost nothing. A video with 10,000 views and a 15% completion rate did worse than one with 1,000 views that people finished. Completion rate is the north star — it tells you whether the hook attracted the right viewers, whether the pacing held, and whether the payoff arrived. Consistently low completion means a mismatch somewhere: the hook is pulling in the wrong people, the video is too long for the value it delivers, or the good part comes too late.

The metric most creators ignore is how many people visited your profile or sent a connection request after watching. A reaction is passive appreciation. A profile visit is active interest — someone thought “who is this, and what else do they make?” That is where a relationship starts. If you want the full map of what LinkedIn reports and what each number is good for, we’ve covered it in our guide to LinkedIn post analytics.

End the video with a specific soft CTA. Not “follow me for more” — something like “I break down a client conversation like this every Tuesday.” You are giving people a reason to expect a particular kind of value.

And check the analytics again a week after posting, then at two weeks. LinkedIn will keep resurfacing content that continues to earn engagement, so a video with legs is worth a second push — share it into a relevant group, reference it in a new post, add a comment with an updated thought.

When someone does land on your profile after watching, make sure it converts. If the video was about sales frameworks, your headline and About section should immediately signal that you are worth following for sales insight. The ultimate guide to personal branding on LinkedIn covers how to align the two.

One Video, a Week of Posts

Video takes more effort than text, and the return on that effort is not one post. It is a text post built on the argument, a carousel of the key points, a still frame with the strongest line, a clip of the best 30 seconds, and material for a newsletter or a blog. Each version reaches people who would never have watched the original.

That is the work ContentIn does — taking the points you already made on camera and turning them into posts written in your own voice rather than a generic one, then keeping the history so you can see which framing actually worked. Not because text beats video; our own benchmarks say the opposite. Because one recording contains more than one post, and the second, third and fourth are nearly free. Our free LinkedIn tools are a reasonable place to start, and you can see what’s in the paid plans on the pricing page.

Start with one video this week. Not a polished production — a clear idea, delivered to camera. Set the thumbnail yourself. Write the caption as a pre-hook. Watch the retention curve. Then do it again next week, better informed about what your audience actually wants.