LinkedIn analytics ·

Dwell time on LinkedIn: the invisible metric that decides your reach in 2026

Dwell time is the time a user actually spends on your post in the feed before carrying on scrolling. LinkedIn has measured it since 2020, introduced it as a key algorithmic signal in late 2022 and in 2026 it is the factor that weighs most in a post's distribution, above likes, short comments and shares with no comment.

This article explains what dwell time measures exactly, why the algorithm prefers it, how the signal is calculated in the first two hours and which concrete writing and formatting decisions raise or sink it.

By Sheena de PunkVoice · Edited by Mario Pérez

Warm illustration of a woman with a side braid in a navy jumper watching a large ornate terracotta analogue stopwatch resting on an open book on a wooden desk, holding it delicately between two fingers, beside a terracotta lamp, mug, plant and a laptop with a blue-lit screen.

What dwell time really is and what it is not

Dwell time is the time, in seconds, the post stays visible on the user's screen while the user is active (not in the background, not with the app minimised). It is not the time from opening the app to closing it, nor the time from starting to read to hitting like. It is a very specific and very short window of attention measured in milliseconds.

LinkedIn separates raw dwell time (total seconds on the post) from dwell time per unit (seconds per 100 characters, or per slide in a carousel). A long post with high raw dwell time but low per unit reads as heavy text people are skimming; a short post with high per unit dwell time reads as dense material worth rereading. Both pay off, but through different routes.

Why it weighs more than likes in 2026

A like is a one second action. A one word comment is a three second action. Dwell time is a continuous signal of real attention, much harder to fake and much more correlated with the content's perceived usefulness.

Engagement pods, coordination between profiles and quick replies from friends manipulate likes and short comments easily. None of those techniques moves a neutral user's dwell time, which is why LinkedIn has given dwell time growing weight since 2023 as a counterweight to the manipulable signals.

In 2026 the feed's internal signal hierarchy runs roughly: dwell time by professional segment, saves, long comments (over 15 words), shares with a comment, the author's replies to comments, short comments, likes. A post that breaks its like record but has low dwell time gets distributed less than one with half the likes and high dwell time.

The first two hours and why they decide everything

LinkedIn measures your post's average dwell time during the first two hours after publishing, comparing it with the average for similar posts (same author, same content type, same professional segment). If your post is above the reference median, it enters the next distribution wave, at 5 to 10 times more impressions. If it is below, it stalls in the first wave.

Those two hours depend heavily on the first impression: whoever sees the post in the first twenty minutes is your direct contacts and most active followers, the people who know you best. If that circle does not stay with the post for more than three seconds, the system infers a stranger will stay even less and cuts distribution before it leaves your close network.

That is why the most important decisions sit in the first two sentences and the first image. Not to maximise clicks, but to maximise seconds of attention from your own circle before the algorithm decides whether to distribute further.

Which editorial decisions raise dwell time

There are five concrete levers that consistently move dwell time upwards, and all of them are writing or formatting decisions within any profile's reach.

  • A hook that requires full reading to make sense. A first sentence with a figure, a data point or a counterintuitive claim holds the reader's eye five or six seconds longer than a generic introduction.
  • A visually breathable format. Blocks of one to three lines separated by white space keep the eye active; compact walls make the scroll continue within the first second.
  • Density of information per line. Every sentence has to add something. Filler sentences ("as we all know", "now more than ever") lower dwell time even when the rest is good, because the eye learns to skip.
  • An attached image or carousel that illustrates without duplicating. An image that extends the meaning of the text adds two or three seconds; a decorative image with no connection adds nothing.
  • A close that demands a pause. A closing sentence with a synthesis or a specific question makes the eye stop for another second or two before moving on. A template sign-off ("what do you think?") does not move the metric.

What sinks dwell time without you noticing

The four patterns that show up in posts with low dwell time are equally identifiable and avoidable.

  • Long sentences chained together with no intermediate punctuation. The eye gets lost after two lines and the scroll continues.
  • Symmetrical lists with no criteria behind them. Emoji bullets at the start of every line are visual markers of the feed's centroid that an experienced reader scans without reading.
  • Excessive self reference at the start ("I have been in this industry for 15 years", "as I was telling you in my last post"). The new reader has no context and skips; the recurring reader already knows and skips too.
  • Banner style images with large text duplicating the post. They add visual weight without adding anything new, and the reader skips the reading because they think they already saw the message.

The LinkedIn algorithm in 2026: how distribution works today

Dwell time as a criterion, not a trick

Writing for dwell time is not writing for the algorithm, it is writing for the real reader. The decisions that raise dwell time (a dense hook, a breathable format, density per line, an image that extends, a close with a synthesis) are the same ones that make a post better to read. In 2026 the algorithm has aligned its incentives with those of the serious reader.

Publishing fewer posts, each one designed to hold attention in the first two hours rather than to accumulate cheap reactions, is the editorial discipline that multiplies reach sustainably with no need for tricks. Dwell time thus becomes the best thermometer of real quality in the professional feed.

Frequently asked questions

Can you see dwell time in LinkedIn Analytics?

Not directly. LinkedIn shows impressions, engagement rate and average video watch time, but it does not publish dwell time per post in 2026. The closest indirect signals are the "see more" expansion rate and the reread rate (profile visits from the post). Both correlate strongly with dwell time.

Do video posts have a different dwell time?

Yes. In video the analogous metric is average watch time and the percentage watched of total length. Videos of 30 to 60 seconds with a completion rate above 60% are the equivalent of a text post with high dwell time, and the algorithm treats them similarly.

Does publishing in low traffic hours help dwell time?

Partly, yes. With less competition in the feed, users spend more seconds on each post. But the distribution window is also smaller, so it mainly pays off for niche profiles with a small, highly engaged audience. Generalist profiles usually prefer peak hours even if dwell time per post drops.

Does editing a post after publishing affect dwell time?

Editing in the first few minutes to fix typos has no effect. Editing substantially after an hour resets part of the accumulated signals and usually sinks the metric. If you have to change something, better to delete and republish (losing the comments) than to deeply edit a post that is already being distributed.

Do carousels have better dwell time by design?

On average yes, 30% to 50% above the equivalent plain text, because every slide is a small attention experiment. That is why the format is so rewarded in 2026, and why a weak carousel cover sinks the whole piece even when the rest is excellent.