7 mistakes that kill your LinkedIn reach (data-checked)

You post consistently, you put real effort into your content — and yet your reach is flat. The reason is usually hiding in details nobody has measured. Here are the 7 most costly mistakes, quantified across 516,144 LinkedIn posts.

In short

→ An external link costs -36% median impressions (448 vs. 705 without one) → 4+ hashtags drop likes by -40% compared to no hashtag at all → Opening with a question hurts reach: -28% median likes vs. average → Text-only is the least amplified format: median 11 likes vs. 24 for carousels → Viral posts stack 4 to 6 tactics — single-tactic posts never cross the threshold

Observational study, correlation ≠ causation, possible sampling bias.

Mistake 1: dropping an external link in the body of your post

This is the most counterintuitive mistake. You want to share a useful resource, so you paste a link. The result: your reach takes an immediate hit.

Across 359,000 posts with and without external links, posts containing a link show a median of 448 impressions — against 705 for posts without one. That's -36%. The algorithm amplifies content that keeps people on the platform; a link pointing elsewhere sends the opposite signal.

The fix: drop the link in the first comment, or redirect to your resource in replies after publishing. The full analysis of external links covers the mechanics and the workarounds.

Mistake 2: stacking hashtags

The "3 to 5 hashtags" myth refuses to die. The data kills it.

| Hashtags | Median likes | |---|---| | 0 | 19 | | 1 | 23 (+21%) | | 2–3 | 17 | | 4–6 | 11 (-40%) | | 7+ | 11 |

One hashtag is the measured optimum. From 4 onward, median likes drop to 11 — a 40% fall compared to using none. The theory that hashtags boost discoverability simply doesn't show up in the numbers. The full hashtag analysis goes deeper on why.

Mistake 3: opening with a question

Starting your first line with a question is a writing reflex. It's also a reflex that costs you reach.

Across our corpus, posts that open with a question have a median like count of 18 — against 25 for the overall average. That's -28% fewer likes. By contrast, opening with a number yields a median of 28 (+17% above average).

Questions generate comments, but they sacrifice reach. If visibility is your goal, open with a precise stat or a bold claim. The article on LinkedIn hooks breaks down 12 hook types and their measured effects.

Mistake 4: publishing text only

Text-only is the default format. It's also the one the algorithm amplifies the least.

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

The gap is striking: text-only gets 2.4× fewer median likes than video. And carousels generate 2.3× more median impressions than text-only (Playbook, 438,413 posts). The mechanism is algorithmic — a format that pushes users to swipe through slides or watch a video creates dwell time, which is the signal LinkedIn weights most heavily.

Mistake 5: publishing at the wrong time

There's no single "best time" that works for everyone, but there are clearly bad time slots.

Posts published in the French morning window (7–10 AM CET) show a median of 16 likes — the worst measured slot. The reason: everyone publishes at the same time, the audience is diluted, and initial velocity is low. At noon CET, the median climbs to 25.

The weekend is massively underrated. With 3× less volume than weekdays, the median is actually higher: 20 vs. 17 on weekdays. Less competition, same active audience.

Mistake 6: neglecting the hook

This is the most diffuse and most expensive mistake over time.

In our Playbook covering 438,413 posts, 80% of viral posts in the top 1% have a structured hook within the first 200 characters. And 61% include a piece of data as proof within those same 200 characters. It's not a coincidence — the hook decides whether someone clicks "see more" or keeps scrolling. No click, no dwell time. No dwell time, no amplification.

You don't need a perfect hook on every post. But a post without a clear hook never triggers the mechanism.

Mistake 7: relying on a single tactic

You find an angle that works — vulnerability, shock stat, numbered list — and you repeat it exactly. That's the single-tactic trap.

Posts that break through the virality threshold stack 4 to 6 writing tactics on the same post: data-driven hook, social proof, vulnerability, backed polarization, open CTA. Each tactic in isolation has a limited effect. Stacking creates a multiplier. This is what our analysis of 516,000 LinkedIn posts calls the stacking law.

What to do right now

Before you publish your next post, run through this checklist:

→ Is there an external link in the body? (if yes, move it to the first comment) → How many hashtags? (aim for 1, never more than 3) → What's your first line — a number or a question? → What format — text only or a visual? → How many writing tactics are you stacking?

LinkPost's Post Analyzer runs this check automatically across 300+ factors before you publish. Free, 30 seconds.

Observational study by LinkPost on 516,144 posts (fresh 2026 corpus, 30,843 profiles, 9.6M snapshots). Measured correlations do not prove direct causation. Possible sampling bias. Detailed methodology in the playbook.