Key takeaway: Carousels generate 278% more engagement than video on LinkedIn, despite video views growing 36% year on year. Format is not a neutral choice. It structures audience behaviour, and carousels structurally produce saves and sends, which are the signals LinkedIn's algorithm weights most heavily in 2026.
Social Insider's 2025/2026 analysis of 1.3 million LinkedIn posts produced a finding that runs against the prevailing content advice of the past two years: video views on LinkedIn are up 36% year on year, yet carousels generate 278% more engagement than video. Both things are true simultaneously, and understanding why requires separating growth in consumption from growth in interaction.
Video is being watched more. It is not being engaged with more. Users scroll past, autoplay triggers, a view is counted. The watch did not produce a save, a share, a comment, or a send. LinkedIn's algorithm in 2026 cares about those signals more than passive views, and carousels are structurally better at generating them.
Why Video Growth Does Not Mean Video Outperforms
The video-first content advice of recent years was based on a reasonable heuristic: where attention goes, content strategy should follow. And attention on LinkedIn did shift toward video. But the engagement metric that matters for distribution, the metric that tells the algorithm to show your content to more people, is not views. It is active interaction.
LinkedIn added Saves and Sends as visible analytics metrics in late 2025. This was not a cosmetic update. It was the platform making explicit what its ranking system had been weighting implicitly: content that people save to return to, and content that people send directly to another person, is content that LinkedIn will distribute more aggressively. Both of those actions indicate genuine value, not passive consumption.
Carousels generate save behaviour for a specific structural reason: they are designed to be consumed in a linear sequence, often contain more information than a single reading can absorb, and frequently contain reference material (frameworks, checklists, step-by-step instructions) that users want to return to. The scroll mechanics of a carousel also make swiping forward feel like progress, which creates a completion impulse that flat text posts and short videos do not trigger in the same way.
Format-to-Purpose Matching
The mistake most content teams make is treating format as a production decision rather than a strategic one. They pick video because it is fashionable, or carousels because someone read that they work, without asking what they want the audience to do after consuming the content.
Different formats produce different audience behaviours, and those behaviours have different downstream effects for the creator.
Video is strong for building familiarity and authority at scale. A user who watches a 90-second video of someone explaining a concept will recognise that person and feel more familiar with them than a user who reads the same content as text. For brand building and top-of-funnel awareness, video's reach advantage is real even if its engagement rate is lower.
Carousels are strong for information-dense content where the audience needs to think, not just watch. A carousel breaking down a five-step framework, or comparing five tools side by side, or walking through a case study, gives the reader control of pace. They can spend more time on slide three if it is the most relevant to them. They can go back to slide one after reading the conclusion. This pacing control is what produces saves: the reader identifies the content as something they want to be able to access again.
Text posts are strong for opinion, narrative, and conversation. They generate comments because they are easy to respond to. The friction of pulling out your keyboard to comment on a video is higher than the friction of commenting on a text post you just read.
How the Algorithm Rewards Saves and Sends
LinkedIn's distribution logic, as inferred from performance patterns across the 1.3 million post dataset, treats Saves and Sends as high-quality signals because they are costly for the user. Clicking Like takes half a second. Writing a comment takes 30 seconds. Saving a post requires deciding it has value worth returning to. Sending it to someone else requires deciding someone you know would benefit from seeing it. These are higher-friction, higher-intent actions that the algorithm treats as meaningful quality signals.
When your content generates saves and sends at a higher rate than average content in your category, LinkedIn shows it to more people in the same category. The distribution loop is self-reinforcing: save-worthy content gets distributed to more people, more of whom save it, which extends the distribution further. A carousel that generates strong save behaviour in the first few hours can continue distributing for days in a way that a video that generated a comparable number of passive views will not.
What This Means for Content Strategy
The practical implication is not "stop making video." It is "match format to the behaviour you want the audience to take, and understand how that behaviour feeds back into distribution."
If your goal is awareness and brand familiarity: video is defensible because reach is the primary metric and views are a reasonable proxy. If your goal is to build a following of people who actively engage with your content and come back to it: carousels outperform because saves and sends are the distribution fuel and carousels generate them more reliably.
For most B2B content creators on LinkedIn, the goal is closer to the second than the first. Building an audience of people who find your content genuinely useful enough to save and share is more valuable than building an audience of passive viewers, and carousels are the format most structurally aligned with that goal.
Building Carousel Content That Earns Saves
Save-worthy carousel content shares consistent characteristics. It contains reference material: something the reader will want to come back to, whether a framework, a checklist, a comparison, or a step-by-step process. It delivers value in a format that is hard to absorb in a single pass. It ends with a clear, specific action or takeaway that makes the preceding slides feel like preparation for something the reader can do.
The opening slide determines whether anyone swipes. It needs to make a specific, credible promise about what the carousel contains. Not "5 tips for marketing" but "The exact brief template I use before every AI content request (and why it matters)." The specificity signals that there is real content inside, not generic advice.
Conclusion
The 278% engagement advantage carousels hold over video is not an anomaly. It reflects a structural alignment between the carousel format, the save and send behaviours LinkedIn rewards, and the information-dense content that professional audiences find genuinely useful. Video is growing in consumption and remains valuable for brand awareness. But for creators optimising for algorithmic distribution and genuine audience engagement, carousels are the format that LinkedIn's current reward structure most consistently amplifies. Match format to purpose, then match purpose to the algorithm.


