Most LinkedIn algorithm advice treats reach as a numbers game: post more, post at the right time, get the comments in early. LinkedIn's own engineers have just described a system that rewards something less exciting and far more useful, which is being clear about who you're talking to.
On 9th November 2026, LinkedIn's engineering team published a detailed write-up of Profile Relevance, the system it uses to find posts to show people from accounts they don't follow. It explains how part of the LinkedIn algorithm in 2026 works rather than announcing an overnight change, and it's one of the most specific things LinkedIn has said publicly about how posts reach new people.
Below, we cover what LinkedIn actually published, how the matching works, why generic posts struggle, and three before-and-after examples for hospitality, property and professional services businesses.
What LinkedIn has published this year
LinkedIn has explained two separate pieces of feed engineering in 2026, and a lot of the coverage blurs them together.
March 2026: the feed rebuild
In March, LinkedIn's engineering blog described a rebuild of how the feed finds posts, replacing several separate retrieval systems with a single one built on large language models. Written by Hristo Danchev and published on 12th March 2026, it's the bigger structural change, and it's where most of this year's "LinkedIn rebuilt its algorithm" headlines come from.
November 2026: Profile Relevance
The November post, by Shiyong Lin and 12 co-authors, zooms in on one part of the feed: posts from people you aren't connected to, which LinkedIn calls Unconnected Content. Profile Relevance is the system that finds those posts for you, and LinkedIn says it has launched as a key source of them.
What's confirmed and what isn't
The post confirms:
what the system reads from member profiles and from posts
that it adds country-level location to both
that a post needs some engagement from outside the author's network before it's recommended
the results of LinkedIn's online tests, including an 18.8% drop in hidden posts
It doesn't say how much of anyone's feed now comes from Profile Relevance, and it gives no direct advice to people creating content. We've stuck to what LinkedIn has said, and where we're drawing our own conclusions, we say so.
How the LinkedIn algorithm decides who sees your post
Your post can reach people in two broad ways, and they work differently.
People who already know you
Your connections and followers are the first audience for most posts. LinkedIn describes network proximity and engagement as the signals traditional recommendation systems rely on to keep the feed relevant at scale, and they still do much of the work here.
People who don't follow you
This is where Profile Relevance sits. LinkedIn's write-up acknowledges that relying on network and engagement makes content outside someone's network harder to discover, particularly for people who are still building an audience. Profile Relevance is designed to close that gap by matching posts to members based on professional context rather than who already knows whom.
Where engagement fits
Most of the coverage has missed this. LinkedIn found that posts matched on topic alone struggled to compete in the feed with posts that had social proof. So it added a minimum threshold: a post must receive at least one like, comment or share from someone outside the author's network before Profile Relevance will recommend it.
Relevance gets your post considered. Early engagement from beyond your connections is what lets this system put it in front of more of the right people.
It matches posts to the reader, not just the writer
This is the part that should change how you write. Profile Relevance builds a summary of each member from the professional information on their public profile:
their headline, summary and key skills
their work history, including companies, job titles, how long they were there and role descriptions
their education, including schools, subjects and degrees
LinkedIn says it doesn't include personal attributes such as a member's name or gender. Its final model pulls the top topics from each profile and turns them into a form that can be compared with posts.
For posts, it reads the text, titles and transcripts, along with author details such as location. It also adds the country from both the reader's and the author's profile, because LinkedIn found its early versions were losing geographic cues. It then looks for posts whose content fits the reader's professional world.
To check whether the matches were any good, LinkedIn's team judged each recommendation against seven factors: topic, location, job title, company, industry, seniority and education. Those are evaluation criteria rather than a published formula, but they give a clear picture of what LinkedIn means by relevant.
Our reading of this is simple. If your buyers are office managers, landlords or finance directors, their profiles are full of the language of those jobs. A post that names that role, that sector and that problem has far more in common with the people you want to reach than one written for "business owners" in general.
Why generic posts cost you LinkedIn impressions
LinkedIn didn't only measure whether recommendations were relevant. It also measured how many were generic, meaning posts that didn't match any professional attribute of the person seeing them, and reported that Profile Relevance reduced that share compared with the system it replaced.
The online results point the same way. In LinkedIn's experiments, feed updates viewed rose 0.54% and the rate of likes, comments and shares rose 0.94%. The bigger figure was an 18.8% drop in people hiding posts. Better matching didn't dramatically change how much people scrolled, it took noise out of their feeds.
For members with little activity history, the effect was stronger, with a 17.7% rise in engagement and a 1.8% lift in daily active users, matched from their profile alone. Someone who has only just joined LinkedIn or rarely logs in can still be shown your post if it fits their professional context.
For anyone posting to win work, the takeaway is uncomfortable but useful. A post that could be for anyone isn't just ignored, it's harder for the system to place in front of anyone in particular.
LinkedIn isn't asking whether your post is good. It's asking who it's for.
Generic vs specific: three rewrites
These are illustrative examples rather than client posts. Each one starts with the kind of post any agency could write, then shows a version a specific reader would recognise as being for them.
Hospitality: a restaurant with a private dining room
Generic: "Planning your Christmas party? We've got you covered. Get in touch to book your table."
Specific: "If you're an office manager or PA booking a Christmas lunch for 20 to 40 people, the venue usually isn't the hard part. It's the dietary requirements that arrive the day before. Our private dining room seats 40, and every booking comes with a pre-order form so allergies and preferences are confirmed a week ahead. Weekday lunches in December still have space."
What changed: The second post names the person who makes the booking, the size of the group and a problem they've actually had. Those are words that appear on the profiles of the people making these bookings, and they give the reader a reason to stop.
Property: a lettings agency
Generic: "The rental market keeps changing. Make sure you're working with an agent who knows what they're doing."
Specific: "If you're a landlord with two or three properties and you manage them yourself, the thing that slips usually isn't finding tenants. It's the renewal dates: gas safety certificates, deposit paperwork and tenancy renewals, all landing at different times of year. We've started giving self-managing landlords a one-page renewals calendar for each property, whether or not they use us. Ask and we'll send one over."
What changed: It speaks to a specific kind of landlord, at a specific scale, about an admin problem they'll recognise. The generic version could be posted by any agency in the country, and there's no professional context in it for the system to match.
Professional services: an accountancy firm
Generic: "Don't leave your tax return until the last minute. Get organised early."
Specific: "If you're a sole trader filing your first Self Assessment, 31 January isn't the only thing that catches people out. If your tax bill is over £1,000, you'll usually need to make a payment on account towards next year's bill at the same time, which can add half as much again to what you pay in January. Better to know now than in the last week of January."
What changed: One reader, a first-time sole trader. One concrete risk, with a real figure. One reason to act now. It's the kind of post people save and forward to someone else.
What to change this month
None of this needs a new strategy overnight. These are the changes we'd make first.
Pick two or three topics tied to the sectors you serve. Staying on the same subjects gives the people you want to reach repeated chances to see you, and keeps your posts recognisable.
Name the reader early. Use the job titles and industries your buyers would have on their own profiles, ideally in the first line or two.
Check your profile location. Profile Relevance adds the author's country to posts, so make sure your profile reflects the market you sell in.
Say the specifics out loud. If you post video, transcripts are part of what LinkedIn reads, so name the reader and the problem in what you say, not just in the caption.
Start real conversations with your existing network. A post needs engagement from someone outside your connections before Profile Relevance will recommend it. Our view is that genuine comments from your network are what carry a post to their wider networks first, which makes that threshold easier to reach. Engagement pods and "comment YES" prompts aren't the answer. They bring in engagement from people who have no interest in what you do.
Keep your headline and experience in line with what you post about. LinkedIn's post doesn't say the author's headline affects distribution under this system, so treat this as good practice for credibility with readers rather than a ranking trick.
Judge posts by who engaged, not just how many. A post seen by 400 of the right people is worth more than one seen by 4,000 of the wrong ones.
What we wouldn't do is throw out what's already working because of one engineering post. If a format or topic is bringing in enquiries, keep it, and use this as a reason to make it more specific.
LinkedIn algorithm FAQs
Did the LinkedIn algorithm change in November 2026?
On 9th November 2026, LinkedIn published an explanation of Profile Relevance, a system that recommends posts from people you don't follow based on your professional profile. LinkedIn says it has launched, but the post explains how it works rather than announcing a sudden switch.
Does LinkedIn show my posts to people who don't follow me?
Yes. LinkedIn calls these posts Unconnected Content, and Profile Relevance is one of the systems that finds them by matching posts to members whose professional profiles fit the content.
Does a post need engagement before LinkedIn shows it to people who don't follow me?
For Profile Relevance, yes. A post must receive at least one like, comment or share from someone outside the author's network before the system will recommend it.
Is Profile Relevance the same as 360Brew?
No. Profile Relevance is a specific system for finding relevant posts from outside your network, described in LinkedIn's November 2026 engineering post. 360Brew is a separate LinkedIn research model that has been widely reported alongside this year's feed changes, and it isn't mentioned in the Profile Relevance write-up.
Why have my LinkedIn impressions dropped?
There's no single cause, but LinkedIn's recent work focuses on showing people fewer irrelevant posts, and its Profile Relevance tests cut hidden posts by 18.8%. If your posts are broad enough to apply to anyone, they're harder to match to the people LinkedIn is trying to serve.
Does my LinkedIn profile affect who sees my posts?
Your profile shapes what you're shown, because Profile Relevance matches posts to the professional information on each reader's profile. On the author side, LinkedIn's post only names details such as location, so keep your profile country accurate and treat the rest of your profile as a credibility signal for the people reading.
Reaching the right people
Reaching the right people on LinkedIn has always depended on being clear about who you're for. What's new is that LinkedIn has described a system built around that idea, and one that measures how often it shows people something generic.
If you want your LinkedIn content built around the clients you actually want to win, then get in touch and we'll take a look at what you're posting now.
Sources: Profile Relevance: Boosting Feed Relevance with LLMs at LinkedIn, LinkedIn Engineering, 9th November 2026. Engineering the next generation of LinkedIn's Feed, LinkedIn Engineering, 12 March 2026.
