LinkedIn For You Feed Test: What It Means at 500 Followers
What the LinkedIn For You feed test confirms, what it leaves open, and a posting plan for small accounts that works whichever tab wins.
By HireSteward team · · 8 min read
In this article
LinkedIn is testing a feed split into two tabs, "Network" and "For You", and only a limited group of members can see it. If you have around 500 followers, the result could decide whether your posts mostly reach people who already know you or also reach strangers who care about your topic. No one outside LinkedIn knows which way this goes, so you need a plan that works either way.
What LinkedIn has confirmed, and what it hasn't
Start with what comes from LinkedIn itself.
- The test exists and has an owner. Chief Product Officer Hari Srinivasan announced it in a LinkedIn post on September 18, 2026, as Net Influencer reported.
- He described the Network tab as one that "helps you stay connected to your network, ie the people you have explicitly connected or followed." He described For You as a way to "stay informed" with perspectives from "real practitioners inside and outside your network."
- Only a limited number of members have the two-tab design, Israel Hayom reports.
LinkedIn has not disclosed several things. It has not said how many people are in the test, how long it will run, or what would trigger a release to everyone (ALM Corp). It also has not said which tab it expects to drive more engagement.
The history argues for caution. LinkedIn tested several alternative feeds last year and topic-based Suggested Feeds in June 2026. Social Media Today notes that neither got a wider launch, and it thinks the Network tab is unlikely to catch on. Treat any confident prediction about the outcome as a guess.
Sources also disagree about what is being tested. Influent reads it as a new For You feed, with the old network feed split off into its own tab. It argues that if this sticks, LinkedIn moves from a network graph toward an interest graph, with a higher ceiling on organic distribution. Social Media Today and ALM Corp describe Network as the new option sitting beside the main feed.
The difference matters. In the first reading, recommendations become the default view. In the second, most members keep the feed they already use, and a minority opts into a connections-only view.
A note on vocabulary: "interest graph" is industry shorthand, not an official LinkedIn feature name. LinkedIn's own description of the feed is "people you know, talking about things you care about."
Topic matching is already in the feed you use today
The tabs are new. The system that matches posts to people by topic is not.
In March 2026, LinkedIn changed how it picks candidate posts for the feed, Search Engine Land reported. It used to draw from separate sources: network activity, trending posts, collaborative filtering and topic-based systems. Those were replaced by a single retrieval model built on LLM-generated embeddings. A transformer-based model then ranks the retrieved posts. It looks at patterns across a member's past likes, comments and dwell time instead of scoring each post on its own.
LinkedIn's own example, reported by Social Media Today, is a member interested in electrical engineering who engages heavily with posts about small modular reactors. The model can match that member across both topics because it understands they are related.
The engineering write-up by Hristo Danchev, published March 12, 2026, adds a detail that matters for new members. The embeddings can infer a member's likely interests from profile data alone, before that member has much activity.
Citing LinkedIn, Israel Hayom says ranking weighs a member's industry, skills and professional experience, plus the content they read, engage with or scroll past. The two-tab design separates network content from recommended content. It does not change how posts are judged.
For a small account, this is the practical point. The system reads what your post is about and looks for members whose history suggests they care about that subject. Your follower count is one input among several.
How much of the feed already comes from outside a member's network is unclear. LinkedIn consultant Melanie Goodman estimates the average feed breaks down like this:
- roughly 31% from first-degree connections
- 25% from second- and third-degree connections
- around 10% from Suggested Posts
Her page does not state a methodology, so read these as one practitioner's estimate, not LinkedIn data.
Two scenarios for a 500-follower account
If For You becomes the default
Distribution would depend mostly on topic fit. Search Engine Land's reading of the March system is that posts showing expertise, and fitting current professional conversations, may travel farther even without existing connections. That is the upside for small accounts.
The main risk is topic drift. Say Monday's post is about hiring, Wednesday's is about a conference and Friday's is about productivity apps. Each post might still find some readers, but your account never builds a consistent signal. A reader who liked one post will also find nothing related on your profile.
There is a trade-off. Strangers reached by topic have no history with you, and many will read one post and move on. Expect this kind of reach to be easier to get, but worth less per impression than reach from people who chose to follow you.
If Network gains traction
Srinivasan's description of the Network tab centers on people you "explicitly connected or followed." If members spend more time there, your 500 followers become your main distribution. Who those followers are matters more than how many there are.
The work shifts toward relationships:
- replying to comments in full sentences
- commenting on your connections' posts
- connecting with people you have actually worked with, or who post in your field
Net Influencer places the test alongside LinkedIn's moves against engagement pods and AI-generated content. Those moves include an option, added earlier in 2026, to report AI-generated posts. A connections-only tab leaves less room for pods and generic AI posts.
| For You wins | Network wins | |
|---|---|---|
| Who sees most of your posts | Members matched by topic, many outside your network | People who connected with or follow you |
| What drives distribution | How clearly a post signals its topic, plus reader behavior such as dwell time | Your relationships and conversations with connections |
| Biggest risk for a small account | Scattered topics that never build a signal | Followers who do not care about your work |
| What to do more of | Narrow topics, concrete posts, commenting in your field | Replying, engaging with connections, choosing connections deliberately |
A posting plan that works either way
The two scenarios reward different things, but both favor a narrow set of topics and posts with real substance. Here are steps you can take this week.
- Pick one to three closely related topics you actually work on. Test each one by listing projects, decisions and mistakes you could write about. If the list runs dry quickly, the topic is too thin or too far from your real work. Goodman also argues that an account with 8,000 focused followers can outperform one with 80,000 unfocused followers. She gives no methodology for that either, but the direction matches how topic retrieval is described.
- Start a notes file. When something at work is worth writing about, write down what happened, what you decided and why. Examples: a migration that went sideways, a pricing call, a client question you keep hearing. Posts built from notes like these are specific from the start.
- Make each post easy for the model to classify. Stick to one topic per post. Name it near the top, using the words practitioners in your field use. Mention concrete things like the tool, the contract term or the failure, instead of abstractions. This also keeps human readers reading, which matters because dwell time feeds ranking.
- Comment on other people's posts in your topics. Your engagement history is one of the signals the ranking model reads. A useful comment also puts your name in front of people who already care about the subject.
- Look after the followers you have. Reply to every substantive comment. If a follow-up question deserves a full answer, turn it into a post. Comment on your connections' posts, not only on posts from big accounts.
- Drop the tactics LinkedIn is pushing back on. That means no pods and no generic AI drafts posted without editing. Members can now report AI-generated posts.
This plan has a cost. When you narrow your topics, some followers who came for your general commentary will lose interest. Writing fewer, more specific posts can also feel slow at first.
If LinkedIn shelves the test like the earlier ones, none of this work is wasted. The plan is built for the March 2026 ranking system, which is already live.
How HireSteward handles this
HireSteward is an AI autopilot for LinkedIn and X built around the notes-first approach above.
- You keep a bank of notes about real work, and the writer builds posts only from those notes. It never invents anecdotes.
- A voice profile learns your tone from your past posts.
- Drafts are checked for AI-sounding phrasing before they reach you.
- You can review and edit every post before it is scheduled and published.
- It can add diagrams, multi-slide carousels published as LinkedIn document posts, and narrated short explainer videos.
It does not do the relationship work. Replying to comments and commenting on other people's posts are still up to you.
The Free plan needs no credit card and allows 3 posts per month. Paid plans are Starter at $49, Pro at $149 and Authority at $349 per month. You can start free.
What to check over the next few weeks
- See whether your own feed shows the two tabs. If it does, spend time in each one. Note what kinds of posts For You shows someone with your profile.
- Watch for news on expansion. Follow Srinivasan's posts and LinkedIn's announcements. Until LinkedIn says otherwise, the test's size and duration remain undisclosed.
- Track your own posts by topic. Log each post in a spreadsheet with its topic, impressions and comments, then compare on-topic posts against everything else. If on-topic posts consistently do better, cut the rest. If they don't, rewrite your opening lines so they name the topic in the words your readers would search for.
Frequently asked questions
- Is the LinkedIn For You feed available to everyone?
- No. Only a limited number of members have the two-tab design with Network and For You. LinkedIn has not said how many people are in the test, how long it will run, or when it might expand.
- What is the difference between the LinkedIn Network tab and the For You tab?
- The Network tab shows posts from people you have connected with or chosen to follow. The For You tab recommends professional content, including posts from people outside your network. LinkedIn's product chief described them as staying connected versus staying informed.
- Can a small LinkedIn account reach people outside its network?
- Yes, and that was already possible before this test. The ranking system LinkedIn launched in March 2026 matches posts to members by topic using LLM embeddings, so follower count is only one input. Posts that clearly signal one professional topic give that system the most to work with.
- Should I change my LinkedIn strategy because of this test?
- Not drastically. A narrow set of topics backed by real work helps if For You wins. Replying to comments and engaging with your connections helps if Network wins. Doing both covers either outcome.
- Will LinkedIn roll out the For You feed to all users?
- It is not certain. LinkedIn tested several alternative feeds last year and topic-based Suggested Feeds in June 2026, and neither got a wider launch. LinkedIn has not announced what conditions would lead to a full release.
Sources
- LinkedIn Trials a 'Network' Tab in Bid to Revive Engagement Between Connections (Net Influencer)
- LinkedIn tests new feed inspired by TikTok (Israel Hayom)
- LinkedIn Tries a Feed Just for the People You Actually Know (ALM Corp)
- LinkedIn tests Network tab to encourage engagement among connections (Social Media Today)
- LinkedIn algorithm and platform updates in 2026 (Influent)
- LinkedIn Interest Graph Explained: Feed Beyond Connections (Commenti)
- LinkedIn updates feed algorithm with LLM-powered ranking and retrieval (Search Engine Land)
- LinkedIn updates its Feed algorithm (Social Media Today)
- LinkedIn Deploys LLM-Based Feed Ranking System To Surface Content Beyond Members' Networks (Net Influencer)
- How to Get the Highest Reach on Your Posts in 2026 (Melanie Goodman, Substack)
