LinkedIn quietly shipped a dedicated vertical video tab in late 2025 and spent the first half of 2026 pushing native video impressions into the main feed at 3-5x the rate of text posts in the same dwell-time band. We ghostwrite for 40+ ecommerce founders. We tracked the shift across 1,200 posts. The pattern is now clean enough to act on — but only if you understand where video actually wins and where it still loses to a tight text post.
The 3-5x Reach Lift Is Real — But Conditional
In Q1 2026, we ran 184 paired tests: same founder, same topic, posted within 7 days of each other — once as text, once as a 35-60 second vertical native video. The aggregated lift:
- Impressions: vertical video averaged 3.4x the text version
- Unique viewers: 2.8x
- Profile views per post: 1.6x (lower than impressions multiple — more on this below)
- Comments: 0.7x — video gets fewer comments per impression, full stop
- Inbound DMs: roughly equal per profile view
So the lift is in top-of-funnel reach, not engagement quality. That's the trade.
Three conditions had to be met for video to win:
- Vertical 9:16 format — horizontal video underperformed text in 78% of tests
- Native upload — YouTube/Vimeo links lost to text every time
- Under 90 seconds — 35-60s was the sweet spot, with completion rate as the dominant signal
If any of those three failed, the video lost to a well-built text post.
The 7-Second Hold Metric Is The New Dwell Time
LinkedIn's video ranking is now keyed almost entirely on 7-second hold rate — what percentage of impressions watch past 7 seconds. We see the cutoff cleanly in analytics:
- Below 38% 7-sec hold: reach collapses after first 90 minutes
- 38-55% hold: distribution holds for 6-12 hours
- Above 55% hold: distribution keeps re-firing for 48-72 hours
This mirrors what TikTok and Reels did 3 years ago. The implication for ecommerce founders: your first 6 seconds are the entire game. Brand intro, logo card, "hey guys" — all of it kills the post before LinkedIn ever decides to distribute it.
The four hook patterns hitting 55%+ hold for our founders:
- Number-led: "I spent $47K on this and learned three things"
- Contradiction: "Everyone says you need a DTC site before Amazon. They're wrong."
- Specific scene: "This is the warehouse where we lost $180K last quarter"
- On-camera question: "Do you know why most Amazon brands plateau at $2M?"
What kills hold rate every time: opening on a face talking, opening with "today I want to talk about," opening with B-roll before any audio.
The 4 Video Formats That Work For Ecommerce Founders
After 1,200 posts, four formats account for ~85% of the wins:
1. Talking head, walking, B-roll cut every 4-6 seconds
The dominant format. Founder filmed on phone, vertical, ambient setting (warehouse, office, trade show, kitchen). Cuts to product B-roll, screenshots, or text overlays every 4-6 seconds. Captions burned in. Avg 7-sec hold: 52%.
2. Screen recording walkthrough with VO
Loom-style recording of a dashboard, Amazon listing, ad campaign, P&L sheet. Founder narrating over it. Works because it gives the viewer something to look at while the voice carries the value. Avg 7-sec hold: 47%.
3. Whiteboard / written-on-camera
Founder draws a concept on paper or a small whiteboard. Visual reveal is the hook. Avg 7-sec hold: 58% — highest in our dataset. Underused.
4. Customer / employee interview clip
15-30 second clip of someone other than the founder — a customer, a team member, a supplier. Counter-intuitively strong on profile views because the founder appears credible by association. Avg 7-sec hold: 44%.
What doesn't work: tutorial-style videos longer than 90 seconds, "podcast clip" repurposes (low completion), motion-graphic explainers with no human face.
When Text Still Beats Video
We don't recommend video for every post. Text still wins in three scenarios:
1. Dense data / framework posts. A 9-point breakdown reads faster than it watches. Saves dominate, and saves still carry the heaviest distribution weight in 2026. Video can't compete on save rate.
2. Controversy / strong opinion posts. Text comment threads are where opinions actually fight. Video comments stay polite and surface-level — we measured a 62% drop in argumentative-vs-supportive comment ratio on video vs text.
3. Long-form storytelling over ~400 words. LinkedIn's video duration cap punishes anything that needs more than 90 seconds of context.
The Cadence Rule: 1 Video Per Week, Not Per Day
We tested daily video for 8 founders over 6 weeks. Reach collapsed for everyone by week 3. LinkedIn appears to throttle creators who flood with video — the algorithm seems to be protecting overall feed mix.
The cadence that holds:
- 1 native vertical video per week (Tuesday or Thursday, AM Eastern)
- 3-4 text posts per week around it
- 1 document/carousel post per week as the third format
Founders who hit this mix saw 2.1x profile views per week vs text-only baselines, with no algorithmic decay over 12 weeks.
The Profile-View Conversion Gap
Here's the catch nobody talks about. Video generates 3.4x impressions but only 1.6x profile views. The conversion rate from impression to profile view is roughly 45% of what text delivers.
Why: video viewers consume passively. They don't tap your name the way they tap a name attached to a written argument they want to dig into. So while video wins on top-of-funnel volume, text still does most of the pipeline work for ghostwritten founder accounts targeting sub-200 ICP companies.
The play: use video to inflate weekly impressions and feed the algorithm "active creator" signal, then ride that signal into your text posts where the actual DMs come from. The two formats compound each other. Neither alone is the answer.
FAQ
Do I need to be on camera for LinkedIn video to work in 2026?
Yes for most founders. The 4 formats that work — talking head, walkthrough VO, whiteboard, interview — all require a recognizable human voice or face. AI-generated avatars consistently underperformed in our tests, with 7-sec hold rates 18-24 points lower than the same founder on camera.
How long should LinkedIn videos be in 2026?
35-60 seconds is the sweet spot. Under 30 seconds doesn't give the algorithm enough completion signal to reward; over 90 seconds collapses completion rate. The hold metric LinkedIn weights most heavily is 7-second retention, not full-video views.
Should I post the same video to LinkedIn and Instagram Reels / TikTok?
You can — but re-upload natively to each platform rather than cross-posting links. LinkedIn's algorithm flags videos that show external platform watermarks (TikTok logo, IG handle in corner), and we've seen 30-40% reach suppression on those posts.
We ghostwrite for ecommerce founders building inbound pipeline on LinkedIn. If video feels like the obvious next move but the production lift feels heavy, that's usually because the format is wrong, not the founder. Get in touch — we'll audit the last 30 days of your content and tell you whether video makes sense for your goal, or whether tight text posts get you there faster.