LinkedIn Video in the US: What the 2026 Data Actually Shows
Contents
- Where the US actually stands on LinkedIn video
- A number that contradicts itself, and what it actually means
- The format hierarchy that has nothing to do with France
- The C.L.I.P. framework: four mechanics that move the needle
- Three video formats worth building around in 2026
- Where video fits in a weekly LinkedIn rhythm
- Four mistakes that quietly kill a video's reach
- Measuring what the view count won't tell you
- What I don't know
- FAQ
- Sources
LinkedIn's own numbers say video on the platform is being watched 36% more year over year. An independently compiled sample of 1.3 million business posts found close to the opposite trend at the account level: average video views down roughly 36% across every follower tier measured, from about 2,430 views to 1,380 for accounts above 100,000 followers. Neither figure is wrong. They measure two different things, and the gap between them is the most useful data point in this piece. LinkedIn video strategy, in practical terms, is the set of format, length and distribution choices that decide whether a video gets watched past the first three seconds or scrolled past. This article works from verified 2025-2026 benchmarks for the US and global LinkedIn market, not from a debate about camera reluctance that has no real equivalent here.
In short: LinkedIn is pushing video hard in the US, its largest market at 270 million-plus members, and platform-level metrics back that push, with video watch time reported up 36% year over year. But Social Insider's independent benchmark of 1.3 million business posts still ranks native documents above video on raw engagement rate (7.00% versus 5.90-6.00% in 2026), and average video views per account fell over the same period. The approach that holds up: treat video as a trust-builder that runs alongside documents, not a replacement for them.
Where the US actually stands on LinkedIn video
LinkedIn counts more than 270 million members in the United States, and roughly 35% of US adults now use the platform, according to Hootsuite's 2026 compilation of the company's own figures. The US alone generates about 31% of LinkedIn's total web traffic, more than any other single country. That scale matters for this conversation: whatever LinkedIn tests on video, US accounts see it first, and they see the most of it.
The platform-level trend is genuinely upward. Video creation grew 27% in 2025, per LinkedIn's figures as relayed by Hootsuite, and video uploads posted double-digit growth in the first quarter of 2026, the third straight quarter at that pace. Paid video spend on the platform grew close to 30% year over year in the third quarter of 2026. LinkedIn has also expanded video discovery on desktop: full-screen vertical playback, algorithmically curated recommendations, and video carousels built into search results, according to Search Engine Journal's coverage of the rollout.
None of that describes a market where video struggles to find viewers. It describes a platform actively building infrastructure to keep more of them watching. That fact has to sit at the front of any honest article on LinkedIn video in the US, because it rules out the easiest shortcut available: reusing a French storyline about video underperformance and assuming it applies here. At the platform level, it does not.
A number that contradicts itself, and what it actually means
Here is where the platform-wide story runs into an account-level one, and where the real US conversation should start.
Lakshman Somasundaram, a product leader at LinkedIn, wrote in a 2025 post that video on the platform "is being watched 36% more year over year," with creation growing at twice the rate of other post formats. That is an aggregate, platform-wide figure, and there's no obvious reason to doubt it: more people are posting video, so more total minutes get watched.
Social Insider, an independent analytics vendor, ran a different measurement on a sample of 1.3 million LinkedIn business posts, collected from January 2024 through December 2025 across 16,645 active business pages. Its finding at the account level: average video views fell roughly 36% year over year, across every follower tier it measured. An account with 1,000 to 5,000 followers went from an average of 190 video views to 155. An account above 100,000 followers went from 2,430 to 1,380.
Both figures can be accurate at once. The reconciliation is arithmetic, not mystery, once you separate what each one counts. Total watch time rises when the number of accounts posting video grows faster than the audience watching it does, which matches what LinkedIn itself reported: video creation growing "twice as fast" as other formats. Divide a growing total by a faster-growing number of publishers and the average slice per publisher shrinks even while the total climbs. A US account weighing a video strategy in 2026 should read "watch time is up 36%" as a statement about LinkedIn's infrastructure, not a promise about their own next post.
Neither dataset breaks results out by country. Both are global samples, and I'm flagging that limit rather than pretending the US number exists when it doesn't.
The format hierarchy that has nothing to do with France
The French version of this argument was built on a domestic finding: carousels beating video by a wide margin on a sample of French posts, engagement rate to engagement rate. Whatever explains that specific result, the US market doesn't need to borrow it, because a smaller version of the same hierarchy shows up in the global data too.
| Format | 2025 engagement rate | 2026 engagement rate (Q2) |
|---|---|---|
| Native document | 7.00% | 7.00% |
| Multi-image post | 6.45% | 6.90% |
| Video | 6.00% | 5.90% |
| Single image | 5.30% | 5.20% |
| Poll | 4.20% | 4.50% |
| Text only | 4.50% | 3.95% |
| External link | 3.25% | 3.30% |
Source: Social Insider, LinkedIn Benchmarks 2026, sample of 1.3 million business posts, no country-level breakdown published.
Documents outperform video on this global sample, not by the 278% margin the French benchmark reported, but consistently, across two full years of data, a gap covered at length in our guide to LinkedIn carousels and document posts. That's worth sitting with before a 2026 content calendar is built around video as the default format. LinkedIn's own marketing describes video as generating "5x more engagement," a figure Hootsuite repeats from the platform's materials. Social Insider's independently sampled benchmark doesn't support that multiple. I'd weigh the independent sample more heavily than the platform's own promotional number on this specific point, for the same reason a claim from the company selling the product gets read differently than one from an outside lab.
Video still belongs in the mix. It simply has to earn its place next to documents, the way any format does, regardless of how much work went into producing it.
The C.L.I.P. framework: four mechanics that move the needle
Format-level averages describe a market. A single video still lives or dies on execution, and the execution mechanics that moved the needle in 2019 are, for the most part, the same ones verified again in the 2025-2026 data. I run four checks with clients, built around one word: C.L.I.P.
Caption. About 80% of LinkedIn video is watched with the sound off, per Zebracat's compiled 2026 figures, and 73% of all video views happen on mobile, where autoplay defaults to muted. Captioned videos hold viewers 32% longer than uncaptioned ones in the same dataset. OpusClip, a captioning tool vendor, reports a wider gap between hand-formatted captions and raw auto-captions: 40% to 60% more average watch time. That figure comes from one practitioner's client testing, not a controlled study, and deserves scrutiny before anyone repeats it as fact. Caption anyway. The downside risk is zero.
Length. Videos under 60 seconds retain close to 87% of viewers on average; cutting further, under 30 seconds, produces up to a 200% lift in completion rate, per the same Zebracat compilation. That doesn't mean every video should run 20 seconds. It means the burden of proof rises with every additional second past the one-minute mark, and a 90-second video needs a reason a 40-second version couldn't have delivered.
Identity. Videos with a human face visible in the thumbnail see click-through rates rise 28%. Unlike the French market, where a camera-forward, conversational register was found to sit awkwardly against a more formal professional culture, the American register that performs on LinkedIn is closer to what already performs on YouTube and TikTok: direct, informal, shot without a studio. Camera-forward is not a liability in this market. Generic is.
Push. A single video rarely justifies its own production cost. A podcast episode, a webinar, an internal talk, any recorded hour of real conversation produces more usable material once cut into shorter pieces than it does as a single upload. This is repurposing, not a fresh shoot for every post, and it's the cheapest way to sustain a weekly cadence without burning out whoever is on camera.
Three video formats worth building around in 2026
Short vertical native video, under 60 seconds, captioned, one message. This is the base format. One idea, one clear point of view, published natively rather than linked from YouTube: native uploads see 38% more engagement than external links, per Zebracat's compilation. It doesn't need a script memorized word for word. It needs a structure: the problem, the answer, the catch.
LinkedIn Live, for interviews and panel conversations. Live video generates 4.6 times more comments per viewer than pre-recorded video, according to Zebracat's compiled figures, and it solves a real production problem: a conversational format with a second person in the room is easier to sustain than a solo monologue to a camera, week after week. A monthly Live, cut afterward into three or four short clips, gives a month of native short-form material from one recording.
Repurposed long-form. If a business already produces a podcast, is interviewed on one, or runs regular webinars, that recorded hour produces more usable video than anything shot specifically for LinkedIn, precisely because the conversation already happened. No one has to perform for the camera; they just have to talk.
One format to retire. The scripted, studio-shot corporate video, two to three minutes with motion graphics over a voiceover, draws views out of curiosity and near-zero engagement once the novelty wears off. Its production cost is out of proportion to what it returns, and the data above explains why: length works against it, and a studio set works against the face-forward identity signal that actually drives clicks.
Where video fits in a weekly LinkedIn rhythm
Video posting frequency across Social Insider's sampled accounts rose from about two posts a month in 2024 to about four in 2025. Document frequency rose from roughly one post a month to two. Image frequency rose from five to seven. Everyone is publishing more of everything, which is exactly the competitive pressure behind the falling average-views figure discussed above.
Against a format hierarchy that still favors documents, and reach that's getting harder to earn across every format, a reasonable weekly mix for a US B2B account looks something like this:
| Content type | Suggested frequency | Role |
|---|---|---|
| Native document / carousel | 2x per week | Reach and information density |
| Short vertical video | 1x per week | Trust and recognition |
| Text with image | 1x per week | Point of view, quick reaction |
| LinkedIn Live | 1x per month | Depth, repurposing source |
Documents carry the reach. Video carries something documents can't: a face, a voice, a reason to trust the person behind the account before a call ever happens. A prospect who has seen someone speak, unscripted, for 40 seconds books a meeting more easily than one who has only read their posts. LinkedIn doesn't publish that number, but it's the reason video stays on the calendar even where the engagement-rate table doesn't favor it.
Four mistakes that quietly kill a video's reach
No captions. The single most expensive mistake, in reach terms. With four out of five views happening on mute, an uncaptioned video loses most of its audience before it delivers a single word.
A weak first frame. LinkedIn shows a static preview before playback starts. No visible face, no legible text on that first frame, and the click-to-play rate stays low, which the algorithm reads as low interest regardless of how good the video actually is.
No accompanying text. LinkedIn is not Instagram. The caption below a video gets read before, and sometimes instead of, the video itself. A hundred to two hundred words that set up the point, and end on an open question, do more for engagement than another five seconds of footage would.
Treating every video as a one-off. Filming a new piece for every single post is the fastest way to stop posting video at all. The formats above that hold up over time, Live and repurposed long-form, exist specifically to remove that constraint, and they held up through LinkedIn's broader fall 2026 ranking changes, covered in our full breakdown of that update.
Measuring what the view count won't tell you
Raw view counts are close to meaningless on their own; LinkedIn counts a view after roughly three seconds, which tells you almost nothing about whether anyone actually watched the thing.
Completion rate is a better signal. Videos with a completion rate above 70% are 3.5 times more likely to get algorithmic recommendation, per Zebracat's compiled figures, which means completion isn't just a vanity number, it's an input to distribution itself, alongside signals like saves and sends covered in detail elsewhere.
Comments relative to views tells you whether a video started a conversation or just got glanced at. A video that pulls comments at a noticeably higher rate than the account's average post is doing something the format alone doesn't guarantee.
Profile visits in the 48 hours after publishing. If a video is doing its job as a trust-builder, it should show up as a spike in profile views, not just in the post's own stats.
Direct messages generated. The metric nobody puts in a dashboard by default, and the one that actually correlates with pipeline. A video that prompts a private message from someone who watched it is worth more than one that racks up views and nothing else.
What I don't know
I don't have a US-specific study that isolates format performance the way Social Insider isolated it globally or the French Visibrain study isolated it for France. Every benchmark cited here is global, and I'm treating that as a real limitation, not a footnote to skip past.
I don't know how much of LinkedIn's reported "35% of US adults" figure reflects active, business-relevant use versus a dormant account. Hootsuite's figure is a usage statistic, not an engagement one.
I don't know whether OpusClip's 40% to 60% watch-time gap between formatted and auto captions holds outside the practitioner's own client base; it's the only source I found for that specific number, and it hasn't been independently replicated as far as I can tell.
I don't have a verified, LinkedIn-sourced comparison of Live versus recorded video performance. The 4.6x figure cited here comes from Zebracat's compiled statistics, which blend sources without individually sourcing each one, and I'd treat it as directional rather than precise.
I don't know how paid video's roughly 30% year-over-year growth in Q3 2026 relates to organic video behavior; the figure describes ad spend, not what unpaid accounts should expect from posting.
FAQ
Is LinkedIn video underperforming in the US the way it does in France?
No verified data supports that claim. US-facing platform metrics show video watch time and creation both growing through 2025 and 2026. The one consistent pattern shared with the French market is structural, not cultural: global benchmarks (Social Insider, 1.3 million posts) still rank native documents slightly above video on raw engagement rate, a gap far smaller than the one reported in France.
What's the ideal length for a LinkedIn video in 2026?
Under 60 seconds retains close to 87% of viewers on average, and cutting to under 30 seconds can lift completion rate by up to 200%, per Zebracat's compiled 2026 figures. Longer videos aren't disqualified, but each additional second past one minute needs to justify itself against that baseline.
Do captions actually matter that much on LinkedIn video?
Yes. About 80% of LinkedIn video is watched muted, and 73% of views happen on mobile, where autoplay defaults to silent. Captioned videos hold viewers 32% longer than uncaptioned ones, per Zebracat's compiled data. There's no documented downside to adding them.
Should documents or video be the priority for a B2B LinkedIn account?
Documents, on raw reach and engagement rate: Social Insider's 2026 benchmark puts native documents at 7.00% engagement versus 5.90-6.00% for video. Video still earns a place in the weekly mix because it builds recognition and trust that a document format can't replicate, not because it out-engages documents.
Is LinkedIn Live worth the production effort?
For accounts that can sustain a monthly conversation format, yes. Live video generates 4.6 times more comments per viewer than pre-recorded video, per Zebracat's compiled figures, and a single Live session, cut into short clips afterward, can supply several weeks of native short-form video from one recording.
How often should a B2B account post video on LinkedIn?
Roughly once a week is a defensible baseline against current benchmarks, alongside two document posts and one text post per week, with a monthly Live as a repurposing source. Social Insider's sampled accounts moved from about two video posts a month in 2024 to about four in 2025, so posting weekly keeps pace with, rather than trails, where the market has moved.
Sources
- Social Insider, LinkedIn Benchmarks 2026: engagement rate by format (documents, video, image, text, poll, link), sample of 1.3 million business posts across 16,645 pages, January 2024 to December 2025; average video views by follower tier; posting frequency shift 2024 to 2025.
- Zebracat, "80+ LinkedIn Video Statistics for 2026", updated June 2026: completion rate and algorithmic recommendation, hook retention, mobile viewing share, sound-off viewing share, caption retention, Live versus pre-recorded comments, native versus external link engagement.
- Hootsuite, "LinkedIn Statistics", 2026: US membership figure, US adult usage share, US share of global LinkedIn web traffic, video creation and upload growth 2025-2026, paid video growth Q3 2026.
- Search Engine Journal, "LinkedIn Video Views Up 36%, New Tools & Courses Available": video discovery feature rollout, desktop full-screen vertical playback, search video carousels.
- Lakshman Somasundaram, LinkedIn Pulse, "Up 36% over the last year, video on LinkedIn is booming", 2025: platform-wide video watch time and creation growth figures.
- OpusClip, "LinkedIn Video Caption & Subtitle Best Practices in 2026", 2026: formatted-caption versus auto-caption watch time gap, based on the author's own client testing.
All In: video that earns a place in the feed, not a slot on the calendar
The mechanics that decide whether a video gets watched (captions, length, a real face, a reason to exist) haven't changed in years. What changed is how much competition every format now faces for the same three seconds of attention.
All In is the B2B media that decodes LinkedIn, expert blog, weekly podcast and newsletter for SME leaders and sales directors who want to turn LinkedIn into measurable growth. An original creation by Patrick de Carvalho, on LinkedIn since 2004. Motto: "I Never Lose."
Discover All In and get what matters on LinkedIn, every week.