Is text-only still worth it on LinkedIn in 2026?

Yes, but with adjusted expectations. In Never Boring AI data, text gets 36 reactions on average and 2,430 impressions, vs 3,975 for image. Text-only is faster to produce and still works well for a solopreneur publishing three times a week: image and video are worth the effort only when the visual genuinely adds something.

Is a LinkedIn carousel (document) more effective than a regular image?

No. In the measured data, document (carousel) averages 58 reactions and 3,462 impressions, while image averages 70 reactions and 3,975 impressions. Image leads on both metrics. Carousel takes more time to produce and is only justified when the content genuinely needs multiple slides to unfold.

What Never Boring AI does not do with this data

Never Boring AI doesn't publish formats you don't produce on your behalf. If you don't make video, the tool helps you get the most out of image or text. This data is meant to calibrate expectations, not force a format on you.

How does Never Boring AI collect these LinkedIn stats?

Users who activate results tracking connect their LinkedIn account to Never Boring AI. Stats (impressions, reactions, comments) are pulled from each user's LinkedIn analytics, on posts they published. No personal data is shared between accounts.

LinkedIn post format: what 958 tracked posts actually show

You want to know which LinkedIn post format gets the most engagement. Most advice out there relies on informal observations or LinkedIn's own publications, and LinkedIn has every reason to push video. Here's what the actual numbers look like across 958 posts measured on 5 active LinkedIn accounts from December 2020 to September 2026.

MVMaxence Vroilant·September 22, 2026·5 min read

Never Boring AI analytics: every LinkedIn post with its format, views and reactions.

Methodology: what these numbers measure

The data comes from 958 LinkedIn posts with full stats, on 5 active accounts tracked in Never Boring AI between December 2020 and September 2026 (Never Boring AI internal data, September 2026). Each post is classified by format as declared by LinkedIn: text, image, document (carousel), video. The metrics used are reactions (likes and others), comments and impressions. Reposts are excluded. Impressions are those declared by LinkedIn Analytics, not estimates.

Numbers by format

Here are the averages across the full period: → Video (75 posts): 124 reactions, 31 comments, 3,254 impressions on average. → Image (662 posts): 70 reactions, 20 comments, 3,975 impressions on average. → Document / carousel (89 posts): 58 reactions, 14 comments, 3,462 impressions on average. → Text only (132 posts): 36 reactions, 7 comments, 2,430 impressions on average. Source: linkedin_stats_posts, Never Boring AI, 958 posts measured on 5 active accounts, December 2020 - September 2026.

What averages don't show

Video wins on reactions, but image wins on impressions. An image post is seen by more people on average (3,975 vs 3,254 for video). This is partly explained by 662 image posts vs 75 videos: the 5 tracked accounts publish far more in image format, and their most active audiences are used to it. Video gets more reactions because it creates more engagement when it lands, not necessarily because it gets wider distribution.

Carousel (document) is also more nuanced than it looks. It gets 58 reactions and 3,462 impressions, lower than image on both metrics. Guides presenting carousel as LinkedIn's king format rely on data from 2021-2023, when the algorithm actively boosted it. In 2026, across these 5 accounts, image leads carousel.

What this means in practice

For a solopreneur or freelance publishing alone, the useful hierarchy is: image is the format to master first (volume, impressions, decent reactions, solo-produceable). Video is worth the investment if you make it regularly and your audience responds to it. Carousel justifies its production time only when the content cannot be said in one image or text. Text only remains valid for posts with strong narrative anchoring, where a visual adds nothing.

What Never Boring AI does with this data

When you activate results tracking in Never Boring AI, the tool pulls your LinkedIn stats post by post. You see your own averages by format, not just general benchmarks. If your text-only posts outperform your images, it's your profile that matters, not the 5-account average above. The stats are there to calibrate your strategy on your audience, not to replicate what works in general.

Track your LinkedIn results in Never Boring AI

Results tracking is included in both plans. Connect your LinkedIn account once, and Never Boring AI pulls impressions, reactions and comments for each post. You see what works on your audience, not a general average.

7-day free trial

Frequently asked questions

The questions that keep coming up on this topic, and how we answer them.

Video leads with 124 reactions on average, ahead of image (70), document (58) and text (36). These figures come from 958 posts measured on 5 active accounts between 2020 and 2026 (Never Boring AI internal data).