A LinkedIn endorsement is a one-click confirmation from a 1st-degree connection that you have a skill listed on your profile. Endorsements do one job well and another badly. They help you get found: LinkedIn says they increase the likelihood that you'll be discovered for opportunities related to your skills. As evidence of how well you do the work, they say very little, because an endorsement records only that someone clicked, not what they saw you do.
That is not a criticism of LinkedIn as a network. It is a question of what an endorsement was designed to say, and what people read into it.
What does a LinkedIn endorsement record?
An endorsement records that a connection clicked to confirm you have a skill you listed. LinkedIn's endorsements help page describes them as a way for 1st-degree connections to validate the skills on your profile, shown under each skill. That is all. It does not record what the person saw you do, when, or whether they work in the field themselves. Two endorsements, one from a long-time collaborator and one from someone you met once, look the same on the page.
That design keeps endorsing quick and friendly, which suits a networking product. It also means the signal carries very little information about depth.
Do recruiters look at endorsements?
They see them, and LinkedIn's own framing is about discovery: endorsements help you get found for opportunities related to a skill. What a recruiter cannot do is read an endorsement count as a measure of ability, because nothing in it describes the work.
A recruiter reading endorsements is really reading three things at once: the size of your network, how active it is, and how generous your connections are with a click. None of those is the skill. LinkedIn itself now steers members towards showing where they applied a skill, a change we cover in LinkedIn Skill Assessments ended: what replaced them.
| What an endorsement shows | What it does not show |
|---|---|
| A 1st-degree connection clicked | What they saw you do |
| The skill is on your list | How well you do it, or how recently |
| Your network is active | Whether the endorser works in the field |
| Many people clicked | Whether any of them checked the work |
Why is an endorsement social currency rather than skill evidence?
Endorsements tend to follow relationships. People endorse colleagues, friends and people who endorsed them. Someone with a large, active network will usually collect more endorsements than someone equally skilled who rarely uses the site.
So an endorsement count is better read as a measure of network and activity than of ability. That is still useful information. It just answers a different question from "can this person do the work?".
How can you tell a meaningful endorsement from a reflexive one?
Read the context, not the count. On the profile itself, a few clues separate the two:
- Who endorsed. An endorsement from someone who worked on the same team, in the same field, carries more than one from a distant connection in another industry.
- What else they wrote. A written recommendation that describes a specific project says far more than a click.
- Whether it fits the history. A skill endorsed by colleagues from the job where it was used is more believable than one endorsed by people who never worked with the person.
Even the best of these is still an opinion without the work attached. Treat it as a reason to ask a question, not as the answer.
What would make a peer signal strong?
A confirmation from another person can carry real weight. For that to happen, a few things need to be true:
- The person saw the work. The signal should come from a specific piece of work, not general goodwill.
- The context is recorded. When it happened, what it was about, and who was involved.
- Volume cannot carry it. Ten friends confirming each other should not outweigh one piece of hard evidence.
- Patterns are checked. Reciprocal confirmations inside a small group should be noticed, not rewarded.
How does Lemma handle confirmations from people?
On Lemma, confirmations from other people are one input among several. Each comes from a distinct person, and only three count towards a skill, so a large network cannot inflate a score. Trading confirmations in a ring, where most of your confirmations come from people you confirmed back, is caught by the integrity check, which can only lower a score. How the SCI score works shows the numbers.
Most of the weight sits elsewhere: live sessions rated by the person you worked with, adaptive assessments, and external work such as repositories or design files. All of it is organised in your Skills Passport, with each item traceable to its source. The methodology sets out the integrity rules.
What should you do with endorsements today?
If you hire, treat endorsements as context about someone's network, not as evidence about a skill. Ask for something you can inspect: work samples, an assessment result, or a record of real sessions. Our technical skills assessment template gives you a rubric to score what you see.
If you are building your profile, prune endorsements that no longer reflect your work: the endorsements help page explains how to hide one, opt out of receiving them, or remove one you gave. There is no need to choose between LinkedIn and evidence. Keep LinkedIn for visibility and conversations, and bring evidence that can be checked when someone needs to know what you can do. We set the two side by side in Lemma vs LinkedIn.
ChatGPT
Claude
Perplexity
