Dwell time: the hidden metric that decides your reach

You can optimize your posting time, your hashtags, your frequency — and still flop. What you don't see in your analytics is the signal that matters most: dwell time.

In short

→ Dwell time measures the time spent on your post before scrolling. It's one of LinkedIn's strongest ranking signals. → The longer and more structured your post, the more dwell time it generates — posts of 1,800+ characters show a median of 25 likes, versus 12 for posts under 300 characters. → Carousels and video are the formats that capture the most attention by design: carousels drive 2.3× more median impressions than text alone. → The first 200 characters are critical — they decide whether the reader clicks "see more" or scrolls past. → You can't measure your dwell time directly, but you can engineer it by choosing the right formats and the right length.

What exactly is dwell time?

Dwell time is the time a user spends pausing on your post without scrolling. LinkedIn distinguishes it from a plain impression: you can see a post in your feed without reading it. Dwell time captures the actual stop — the pause.

LinkedIn's Engineering team has documented this signal as part of feed ranking. It's a passive behavioral signal: LinkedIn doesn't need a like or a comment to know you read something. The pause alone is enough to send a positive signal.

The direct consequence: a post someone reads for 45 seconds — with zero reactions — can outperform a post that gets 3 likes from someone who never read it. The algorithm values attention, not surface-level performance.

Why this is the signal you underestimate most

Most creators optimize variables they can see: likes, comments, posting time. Dwell time is invisible in LinkedIn's native analytics.

And yet, it conditions everything else. The algorithm works in two stages: a small sample of your network sees your post first. If that sample pauses and reads, LinkedIn amplifies distribution. If that sample scrolls, the post is buried — regardless of its quality.

That's exactly what our analysis of 438,413 posts confirms: early velocity and time spent on the post are the strongest determinants of distribution. Engineering dwell time means engineering that first sample.

What drives dwell time — the data

Two variables correlate strongly with engagement, and both act directly through dwell time: length and format.

Length: engagement doubles between short and long posts

Across our corpus of 359,000 posts (tracked over 180 days, analyzed in our 516,000-post LinkedIn study), here are median likes by character count:

| Post length | Median likes | |-------------|-------------| | Under 300 characters | 12 | | 300–700 characters | 17 | | 700–1,200 characters | 19 | | 1,200–1,800 characters | 21 | | 1,800+ characters | 25 |

Median engagement doubles between the shortest and longest posts. And posts of 1,500+ characters generate +49% engagement compared to short posts.

The mechanism is straightforward: a longer post takes more time to read. More dwell time, more signal sent to the algorithm, more distribution. For a deep dive into each length bracket, the ideal post length guide covers the full data.

Format: visual formats capture attention by design

Format also acts directly on dwell time — it shapes reading behavior structurally:

| Format | Median likes | |--------|-------------| | Text only | 11 | | Image | 20 | | Carousel | 24 | | Video | 26 |

Carousels force the reader to swipe slide by slide. Video imposes a minimum viewing time. Both are dwell-time generators by construction. In our playbook study of 438,413 posts, carousels deliver 2.3× more median impressions and 4× more comments than text alone.

For a detailed format comparison, the full text vs. image vs. video vs. carousel breakdown has everything you need.

How to engineer your dwell time concretely

You can't open a dashboard and read "dwell time = 38 seconds." But you can build posts that mechanically stop the scroll.

→ Start with a strong hook — the first 200 characters decide whether someone clicks "see more." No click, no dwell time. → Write long and airy — short paragraphs, but many of them. The reader moves forward, stays on the post, keeps reading. → Use carousels for reach — every swipe is a micro-engagement, and LinkedIn records consultation time across slides. → Structure your post with lists, white space, and visual rhythm. A wall of text gets scrolled. A structured post gets read to the end. → Anchor in the concrete — a number, a lived experience, a specific situation. These are the posts readers finish, not the ones they abandon halfway. → Reply fast to early comments — staying active in the first few minutes extends your post's lifespan and drives a second wave of readers.

Before publishing, run your post through LinkPost's free analyzer — it evaluates your first 200 characters, length, format, and 30+ criteria that directly feed engagement and dwell time.

The takeaway

Dwell time is the metric you can't see but that carries the most weight in distribution. You work on it indirectly — by choosing the right formats, writing longer, making every paragraph pull toward the next.

This isn't theory: it's what high-performing creators do differently, visible in the patterns of our 516,000-post LinkedIn study.

Methodological note: the length and format data cited in this article come from the LinkPost observational study (516,144 posts, 9.6M snapshots, 30,843 profiles). The correlations observed (median likes, impressions) do not prove direct causation with dwell time, which is an internal LinkedIn signal not exposed via its APIs. The dwell-time interpretation is grounded in LinkedIn Engineering's public documentation. Correlation does not imply causation, and sampling bias may exist.