> **Methodology in brief.** First-party LinkPost data: **57,809 real LinkedIn posts** and their engagement (likes, comments, impressions), analyzed by format and by text length. Every "virality" figure below is **measured** on this dataset. Third-party industry comparisons are cited and dated. Central honesty point: these are **correlations**, not guarantees — correlation ≠ causation. No invented figures.

## Key takeaways

- **Virality is a statistical exception.** Median sits at **27 likes/post**, mean 115 (a heavily skewed distribution pulled up by a long tail). *(LinkPost, n = 57,809)*
- **Only 17.4% of posts reach 100+ likes** and **3.7% reach 500+**. The p90 is **178 likes**, the p99 is **1,786**, the observed max is **57,688**.
- **Format matters**: video **146 likes** > image **127** > carousel **100** > text **65** on average. Image drives the most **comments** (35). *(LinkPost, n = 57,809)*
- **Length has a double optimum**: short (<300 chars) maximizes **likes** (135); **800–1500 characters** maximizes **comments and impressions**. The trough is in the mid-short range (300–799 chars: 103 likes).
- **"Viral" can't be ordered on demand.** The levers (video/image format, fitting length, and — per our other studies — a strong hook + early engagement) **raise your odds**, they don't guarantee anything.

## 1. LinkedIn virality is rare (the distribution)

Before chasing "the recipe," face the reality. Across **57,809 real posts**, here's how likes are distributed.

| Measure | Value (likes/post) |
|---|---|
| **Median** | **27** |
| Mean | 115 |
| p90 (top 10%) | **178** |
| p99 (top 1%) | **1,786** |
| Observed max | **57,688** |

| "Virality" threshold | Share of posts that reach it |
|---|---|
| **100+ likes** | **17.4%** |
| **500+ likes** | **3.7%** |

**Reading.** The median LinkedIn post gets **27 likes**. The mean (115) is **more than 4x the median**: that's the signature of a heavily skewed distribution, where a small minority of viral posts (up to **57,688 likes**) drags the average up while most posts do modestly. In plain terms: **more than 8 in 10 posts never cross the 100-likes line**, and only about **1 post in 27** clears 500.

The honest lesson: virality is **not** a post's normal state. It's the exception. Everything below raises your **odds** of landing on the good side of the distribution — without ever guaranteeing it. To compare these figures with the reach of a comment, see [comment or post on LinkedIn](/en/blog/commentaire-vs-post-linkedin).

## 2. Format: video and image lead

The first measurable lever is post **format**. Here's average engagement per format across our 57,809 posts.

| Format | Avg. likes | Avg. comments |
|---|---|---|
| **Video** | **146** | ~ |
| **Image** | 127 | **35** *(highest)* |
| Carousel | 100 | ~ |
| Text only | **65** | ~ |

**Reading.** **Video** generates the most likes on average (**146**), followed by **image** (127) and **carousel** (100). Text-only trails (**65**) — roughly **2.2x fewer likes than video**. On the **comments** side, the **image** sparks the most conversation (**35**).

A word of caution on interpretation: these gaps are **correlations**. Video/image posts are often produced by more advanced creators with a more engaged audience — part of the gap comes from **who** posts, not just **what**. Format is not a magic button, but a **lever** that pushes in the right direction. To dig deeper, read our dedicated study on [LinkedIn post format and engagement](/en/blog/format-post-linkedin-engagement).

## 3. Length: a double optimum (likes vs comments)

The second lever is **text length**. And here the answer is neither "shorter is better" nor "longer is better." It depends on what you optimize for.

| Post length | Avg. likes | Optimize for |
|---|---|---|
| **Short (<300 chars)** | **135** | **Maximizes likes** |
| Mid-short (300–799 chars) | **103** *(the trough)* | — |
| **Long-medium (800–1500 chars)** | — | **Maximizes comments + impressions** |
| Long (1500+ chars) | 126 | Keeps strong likes |

**Reading.** **Short posts (<300 characters)** collect the most **likes** (**135**): a punchy, scannable, easy-to-like message. But to trigger **comments** and maximize **impressions**, posts of **800 to 1500 characters** win — long enough to develop an idea and invite debate. The **trough** sits in the mid-short range (**300–799 chars: 103 likes**): too long for the quick-like reflex, too short for substantive debate. Good news for long-form: beyond **1500 characters**, likes stay solid (**126**).

The practical takeaway: **short for likes, long-medium for conversation and reach.** Pick based on your goal for the day.

## 4. So what makes a LinkedIn post go viral?

Let's be clear: **nothing "makes" a post viral deterministically.** The data (median 27 likes) says so bluntly. But our 57,809 posts surface a list of levers that **tilt the odds**:

- **Format** (section 2): video and image beat text-only on average.
- **Length** (section 3): short for likes, 800–1500 chars for comments and impressions.
- **Hook + early engagement**: our other studies show it — the first hour and the first line decide much of the reach. That's exactly what the algorithm watches ([how the LinkedIn algorithm works in 2026](/en/blog/algorithme-linkedin-2026)), and the heart of the [LinkPost Algorithm Playbook 2026](https://www.linkpost.gg/fr/playbooks/linkedin-algorithm-playbook-2026).
- **Debate**: 2026 third-party sources agree that posts sparking conversation get distributed more ([Metricool, 2026 study](https://metricool.com/press-release-linkedin-study-2026/)).

> ⚠️ **Correlation ≠ causation.** All these levers are measured *on average* across existing posts. They **raise your odds** of virality; they don't **cause** it mechanically. A great post can get 27 likes; an average post can explode. The role of randomness is real.

This is exactly the problem **[LinkPost](https://www.linkpost.gg/fr)** solves: instead of publishing blindly and hoping, it **predicts a post's viral potential BEFORE publishing** via a **33-criteria** model (format, hook, length, structure, expected engagement signals). You see the score, you fix, you publish — you play the probabilities instead of being subject to them.

## 5. What third-party studies say (2026)

Our first-party numbers are consistent with the 2026 industry trend:

- **Video/document formats dominate reach**, but "visible" engagement (likes, comments) is declining in favor of "invisible interactions" — carousel swipes, video views, clicks. *([Metricool — 2026 LinkedIn Study](https://metricool.com/press-release-linkedin-study-2026/))*
- **Overall vanity metrics are dropping**: per [Richard van der Blom (Algorithm Insights 2025)](https://www.linkedin.com/posts/richardvanderblom_chapter-1-algorithm-insights-report-2025-activity-7322514599126130688-Q895), views and public engagement are down year over year — which makes the modest median (27 likes) even more credible. *(third-party, dated 2025)*
- **The first hour stays decisive.** Studies converge: early engagement conditions reach expansion. *(industry trend 2025–2026)*

These sources confirm our angle: virality is rare, format and timing matter, and **no** recipe guarantees success. For the full numbers picture, see our [LinkedIn statistics 2026 roundup](/en/blog/statistiques-linkedin-2026).

## FAQ

**What makes a LinkedIn post go viral?**
No single factor guarantees it. Our data (57,809 posts) shows virality is rare — median 27 likes, only 17.4% of posts reach 100+ likes. The levers that **raise your odds**: video/image format, fitting length (short for likes, 800–1500 chars for comments/impressions), a strong hook and early engagement. These are correlations, not guarantees.

**How many likes does an average LinkedIn post get?**
The **median is 27 likes** (mean 115, pulled up by a long tail of viral posts). The p90 is 178 likes, the p99 is 1,786. *(LinkPost, n = 57,809)*

**What percentage of LinkedIn posts go viral?**
Depending on the threshold: **17.4% of posts reach 100+ likes** and **3.7% reach 500+ likes**. The vast majority do modestly. *(LinkPost, n = 57,809)*

**Which post format performs best on LinkedIn?**
On average, **video** (146 likes) leads image (127), carousel (100) and text-only (65). For **comments**, the **image** is on top (35). These are correlations — format is only one lever among several.

**What post length goes viral?**
It depends on the goal: **short (<300 chars)** maximizes **likes** (135), while **800–1500 characters** maximizes **comments and impressions**. The trough is between 300 and 799 characters (103 likes).

**Can you predict whether a post will go viral?**
You can't guarantee it, but you can **estimate its potential** before publishing. That's what [LinkPost](https://www.linkpost.gg/fr) does with a 33-criteria model that scores the post BEFORE publishing it — you play the probabilities instead of posting blindly.

## Sources & methodology

- **LinkPost dataset** — real engagement on **57,809 LinkedIn posts**: likes median **27**, mean **115**, p90 **178**, p99 **1,786**, max **57,688**; **17.4%** ≥ 100 likes, **3.7%** ≥ 500 likes. By format (avg. likes): video 146, image 127, carousel 100, text 65; avg. comments: image 35 (highest). By length (avg. likes): short <300 chars = 135, 300–799 chars = 103, 1500+ chars = 126; 800–1500 chars = optimum for comments + impressions. See [LinkPost](https://www.linkpost.gg/fr) and its [2026 algorithm playbook](https://www.linkpost.gg/fr/playbooks/linkedin-algorithm-playbook-2026).
- **Nature of the figures**: all **measured** on the LinkPost dataset. The levers are **correlations** (correlation ≠ causation) — they raise your odds of virality, they don't guarantee it.
- Third-party sources: [Metricool — 2026 LinkedIn Study](https://metricool.com/press-release-linkedin-study-2026/) · [Richard van der Blom — Algorithm Insights 2025](https://www.linkedin.com/posts/richardvanderblom_chapter-1-algorithm-insights-report-2025-activity-7322514599126130688-Q895).
- Sister studies: [Comment or post on LinkedIn?](/en/blog/commentaire-vs-post-linkedin) · [LinkedIn post format and engagement](/en/blog/format-post-linkedin-engagement)