What is the authenticity score?
The authenticity score shows how genuine the engagement on a creator’s LinkedIn posts is. It looks for signs of fake or bought engagement, such as repetitive comments, the same group of people commenting on everything, or bursts of comments right after posting. It also checks whether the comments carry real conversation.
The score goes from 0 to 100. Every creator starts at 100 and loses points for each sign of artificial engagement found on their recent posts.
Which networks does it apply to?
The authenticity score is currently available on X and LinkedIn only. On other networks, it doesn’t affect your score for now, so nobody gets an advantage or a disadvantage.
Which posts are analysed?
The score is based on your last 10 LinkedIn posts from the past 6 months that meet these conditions:
The post is at least 7 days old, so its engagement has had time to settle.
The post has at least 10 comments from other people. Your own replies don’t count.
On each post, the score reads the most-liked comments and some of the replies under them.
You need at least 5 posts that meet these conditions to get a score.
What does the score look at?
Points are deducted when the analysis finds any of these signals:
1. Copy-paste comments
Comments that look templated or generic, or that are built on the same sentence frame. Comments from people who only commented once count much less, since creators can’t control what strangers write.
2. The same commenters on every post
A large share of your comments coming from the same small group of people. A small share is normal and isn’t counted.
3. Bursts of fast comments
An unusual wave of quick comments from regular commenters shortly after posting. This isn’t counted when the post asks people to comment, or when you reply to many of the comments.
4. Empty comments
Emojis or very short comments that get no likes and no replies. A reasonable share is normal and isn’t counted. Announcement posts are not checked for this, since short “congrats” comments are expected there.
5. Media made by others
Posts that use someone else’s images or videos. Credited media counts less, and credited media in a long opinion post doesn’t count at all. Your own show, podcast, talks, book, product, rankings or certificates count as your own.
6. AI-style writing
Long posts that show strong signs of AI-generated writing, or AI-generated images. This only counts when another signal is also found. Writing with AI tools is never penalised on its own.
If a signal can’t be measured, for example because there aren’t enough long posts, it doesn’t cost any points.
What is audience quality?
The score also looks at the quality of the conversation under your posts, using four grades:
Comment substance: do comments share experiences, ask questions, add information or bring a different view?
Interaction: do comments get likes and replies, are there real discussions, and do you respond meaningfully?
Expertise: does your professional background match the topics you post about, and do people from your field engage?
Post content: do your posts share your own work, ideas and specifics?
When audience quality is very low, the score is held lower even if there’s no sign of fake engagement. It applies to audiences where nothing looks fake, but very little real conversation is happening.
Authenticity score bands
Band | Score |
High | 80 and up |
Good | 60 to 79 |
Fair | 40 to 59 |
Low | under 40 |
Does it affect my global score?
No. Your global score is always neutral on authenticity.
What do brands see?
Brands see:
The creator’s band and what it means
A short summary of where points were lost
A breakdown of each signal: what was found, how it counts, why it matters and what it doesn’t mean
An explanation of each audience quality grade
What do creators see?
Creators see their band and the date of their score. They also see the signals that had a noticeable impact on their score, each with a tip on what they can do.
What the authenticity score does not do
It does not judge follower count, reach or performance.
It does not penalize writing with AI tools on its own.
It does not name the people who left generic comments.
It does not ask creators to delete comments.
A lower score does not automatically mean a creator is fake. It can also reflect an audience that engages less in the comments.
Frequently asked questions
Does having more followers improve my score?
No. The authenticity score doesn’t take follower count, reach or performance into account.
Why is authenticity applied to my LinkedIn score but not my other networks?
The authenticity score is only available on LinkedIn and X for now. On other networks, it doesn’t affect your score, so nobody gets an advantage.
A creator has a Low score. Does that mean they’re fake?
Not necessarily. A lower score can also mean the audience is less active in the comments. Brands can check each part of the score in the breakdown.