LinkedIn Virality Score: predict your post before you publish
LinkPost's virality score rates your LinkedIn draft before you publish on 300+ factors (including 27 personalized to your style), via an algorithm calibrated on 1 million analyzed posts. It adds a controversy score and a sentiment score, and tells you what to fix. It's a statistical prediction, never a guarantee.
TL;DR
- 1You paste or write your draft, LinkPost scores it BEFORE publishing across three axes: virality, controversy, sentiment.
- 2The score draws on 300+ factors analyzed by the V2 algorithm, including 27 personalized to your voice, calibrated on 1M+ LinkedIn posts.
- 3On the narrative side, the marketing grid frames it as 33 criteria (22 attention tactics + 11 virality rules): a more readable view of the same levers.
- 4The algorithm doesn't just give a grade: it pinpoints the weak link (for example, your hook caps at 0.6/1) so you know exactly what to rewrite.
- 5A score is a prediction, not a promise: across the 20 most active users, +71% of posts break through and -75% of flops avoided, but no guaranteed virality.
How the virality score works
You write or paste your draft, and LinkPost runs it through the scoring engine before you publish. The engine doesn't settle for a single grade: it breaks your text down across 300+ factors (hook, structure, mobile readability, idea density, engagement triggers) and returns a numeric score, with the detail of what's dragging the post down.
In practice, you see where it breaks. The algorithm tells you, for example, "your hook caps at 0.6/1" instead of leaving you to guess. You fix the weak link, re-score, and publish the version that matches the most winning patterns. The goal isn't a good grade: it's knowing what to rewrite.
The diagnosis stays concrete, never vague. The score tells you your closing line falls flat and offers two alternatives, or that one posting time loses you 40% of impressions in your niche. It's science replacing intuition.
Three scores, not one
LinkPost doesn't return a single number. It computes a virality score (engagement potential), a controversy score (how much tension the post carries) and a sentiment score (the dominant emotional tone). You steer all three together, because a post can break through for the right reasons as well as the wrong ones.
300+ factors, 27 personalized
The V2 algorithm analyzes 300+ factors, including 27 personalized to the way you write (rhythm, vocabulary, preferred formats). The score isn't a generic grid: it accounts for what works for YOU, learned from your posts, not from an anonymous average.
33 criteria, the readable view
On the surface, LinkPost sums all of this up into 33 criteria: 22 attention tactics (catching the eye, holding the read) and 11 virality rules (what drives sharing). It's not a separate algorithm: it's a human-readable view of the same levers, so you understand the score instead of just receiving it.
A prediction, never a guarantee
Let's be honest: no tool can promise virality. The score is a statistical prediction based on the patterns that correlate with engagement across 1 million analyzed posts. It clearly raises your odds and shows you what to improve, but real performance also depends on timing, your audience, current events and a degree of randomness no one controls.
That's exactly why a good score reads as a signal, not a certainty. Across LinkPost's 20 most active users: +71% of posts break through and -75% of flops avoided. These are measured results, not an individual guarantee on your next post. You put the odds on your side, you don't cheat them.
Aligned with LinkedIn feed research
The score didn't come out of nowhere. It was designed by drawing on LinkedIn Engineering's public work on the feed: a ranking that relies on embeddings to understand a post's meaning, and that distinguishes passive signals (time spent reading) from active signals (like, comment, share).
Put simply, LinkPost doesn't chase vanity metrics: it optimizes for what the platform actually values. The score reflects the mechanics described by LinkedIn's own engineers, not recycled guru recipes.
Free analyzer vs full prediction
There are two levels. The Post Analyzer is free and public, no signup required: you paste a post, you get a score on 300+ factors to test the logic. It's the perfect tool to understand the method before committing.
The full prediction lives in the product: it adds personalization to your voice (27 factors), the controversy and sentiment scores, and the loop of write, score, schedule, publish, measure. There's no free trial of the full product: the Analyzer is the free entry point, and LinkPost is the next step when you want to score and run things over time.
Scoring before you publish vs publishing blind
| Criterion | Publishing blind | LinkPost |
|---|---|---|
| Decision | By instinct, we'll see | Numeric virality score before publishing |
| Axes scored | None, you publish and pray | Virality, controversy, sentiment |
| Depth of analysis | Eyeball re-read | 300+ factors, 27 personalized |
| Diagnosis | No feedback | Weak link pinpointed (e.g. hook at 0.6/1) |
| Reference base | Your gut feeling | 1M+ LinkedIn posts analyzed |
You stop publishing with crossed fingers: you see the score, fix the weak link, and decide knowingly. The score stays a prediction, not a promise.
LinkPost methodology, based on 1 million analyzed LinkedIn posts
Source: Engineering the next generation of LinkedIn's Feed (LinkedIn Engineering)
Frequently asked questions
- Is it 33 criteria or 300+ factors?
- Both describe the same thing at two levels. The V2 algorithm analyzes 300+ technical factors, including 27 personalized to your style. On the surface, LinkPost sums them up into 33 readable criteria (22 attention tactics + 11 virality rules) so you can understand the score.
- What's the difference between the free Analyzer and the full prediction?
- The Post Analyzer is free, public and requires no signup: you paste a post, you get a score on 300+ factors. The full prediction, inside the product, adds personalization to your voice, the controversy and sentiment scores, and the complete write-score-publish-measure loop.
- What is the score calibrated on?
- On a base of more than 1 million analyzed LinkedIn posts, and designed by drawing on LinkedIn Engineering's public research on its feed (embeddings, passive vs active signals). It optimizes for what the platform truly values, not for vanity metrics.
- What do the controversy and sentiment scores measure?
- The controversy score gauges how much tension your post carries (a divisive post engages more, but not always for the right reasons). The sentiment score identifies the dominant emotional tone. You steer all three axes together to publish with full awareness.
- Do I need a free trial to test the score?
- The full product has no free trial, but the LinkedIn Post Analyzer is free and requires no signup. You can test the scoring logic there before moving to the full prediction, personalized to your voice.
About LinkPost
LinkPost is a content tool for LinkedIn, co-founded by Yannis Haismann and Matteo Kocken, two engineers. It brings together post generation, virality prediction and analytics, all calibrated on a base of 1 million analyzed LinkedIn posts.
This feature is part of the full product: it works hand in hand with generating 25 scored variants per topic, the 33-criteria, 300+ factor virality prediction, and your own analytics. The goal is never to publish blind, but to decide with data.
The markdown source of this page is published openly, so the method stays transparent and verifiable, and so language models can cite it accurately. The numbers cited come from first-party measurement, the limits are stated honestly, and every claim links back to a source you can check. There is no free trial of the full product: the Post Analyzer is the free tool, and LinkPost is the next step.
Keep going with LinkPost
Paste your draft, look at the score, fix the weak link, and publish knowingly instead of crossing your fingers.
Score my post