Field Manual LinkedIn

LinkedIn: the feed that counts the seconds

The Major's most decorated front, and his most crowded cemetery. LinkedIn tells you more about its feed than anyone. Listen.

Sources checked 13 October 2026.

Roll call of the fallen

Private Pod joined a group of forty that liked each other's posts every morning at 7:00. His numbers climbed. Then his reach outside his own network quietly shrank, and a warning arrived. Corporal Comment-to-Agree asked for a "YES" in the comments; LinkedIn now names that exact prompt as bait it reduces. Sergeant Great-Share left "Great share!" under two hundred posts with a tool, and his comments vanished from "Most relevant." We salute them. We do not repeat them.

(Composite soldiers, recruit. Every rule they broke is quoted below.)

What LinkedIn rewards

Time spent with your post. LinkedIn's engineers predict how long a member will spend on a post and use a short look as a negative signal. Below a certain time, they found "the probability of engagement is nearly zero," so those views count as skips (LinkedIn Engineering). Their ranking research describes a "long dwell" signal for people who read without clicking anything (LiRank, LinkedIn, 2024). So a post people actually read is doing its job, even when nobody likes it.

Relevance to what people care about at work. In March 2026 LinkedIn said it was rolling out ranking that "better understand[s] what posts are about and how members' interests change over time," using "signals you choose to share, like your industry, skills, experience, and geography, alongside how you engage with content over time" (LinkedIn Pressroom, 12 Mar 2026). Its engineers say the new ranker reads "more than a thousand of your historical interactions" and relies on "professional signals and engagement patterns, never demographic attributes" (LinkedIn Engineering). Posts can reach "professionals you haven't met yet" when they match those interests.

Real conversation. LinkedIn says authentic comments "often can spark meaningful conversations and connections" and that it protects them by keeping automated comments out of "Most Relevant" (LinkedIn Pressroom, 13 Mar 2026).

Your relationship with the reader. LinkedIn's ranking research found that how often a viewer has engaged with an author still matters, even in its newest model (Feed SR, LinkedIn, 2026).

Credibility. LinkedIn is expanding verifications and says more than 100 million members have added at least one (LinkedIn Pressroom, 13 Mar 2026). Hypothesis: a verified profile helps readers trust you; LinkedIn hasn't said it changes ranking.

What LinkedIn cuts, and what gets you benched

Engagement bait. "We're improving our systems to reduce repetitive, low-substance posts and engagement bait, such as 'comment to agree' prompts or videos that don't match the text" (LinkedIn Pressroom, 12 Mar 2026).

Recycled posts. Same source: "We're reducing recycled and click-driven posts... We want the Feed to be relevant to your interests, not a popularity contest."

Engagement pods. A pod is "a coordinated group of people that like, comment and share posts to attempt to boost visibility and virality," and it "is not allowed on LinkedIn." LinkedIn has removed groups that showed pod behavior and warned members of "possible account restrictions and removals from LinkedIn programs, including Top Voices" (LinkedIn Pressroom, 13 Mar 2026). Its policies put it plainly: "don't agree with others ahead of time to like or re-share each other's content" (Professional Community Policies).

Automated comments. Comments posted "using a browser extension, script, or third party tool" without a person clicking "comment" "aren't allowed." LinkedIn may hide them from "Most Relevant," hide them outside the commenter's network, or "restrict a member's ability to use LinkedIn" (same source).

Automation of the website, in general. LinkedIn says it doesn't permit "browser plug-ins, or browser extensions that scrape, modify the appearance of, or automate activity on LinkedIn's website" (Prohibited software and extensions). Its User Agreement section 8.2 bans using "bots or other unauthorized automated methods to... create, comment on, like, share, or re-share posts." Blowhorn is a browser extension. It acts at human pace in your own session, but LinkedIn's rule covers it all the same. That's the honest truth, and the reason the rest of this page is so careful.

Too many invitations. LinkedIn doesn't publish a number. It says you can be restricted when "You've sent many invitations within a short amount of time," when "Many of your invitations have been ignored, left pending, or marked as spam," and that "If you send an excessive number of invitations and we suspect the use of an automation tool, we may suspend or restrict your account" (Types of restrictions). A restriction "typically lasts one week," withdrawing invitations "will not remove the restriction," and support "cannot remove or shorten the wait" (Invitation limit reached). Free members can add a personal note to five invitations a month (same source).

Silent reach cuts. LinkedIn may limit how a post is "distributed across LinkedIn" if it suspects inauthentic engagement (LinkedIn Pressroom, 13 Mar 2026). Nobody gets a letter. Watch your reach.

Do's

  • Write for the reader who stops scrolling. A post someone reads to the end earns a long-dwell signal even without a like (LinkedIn Engineering).
  • Stay in your lane, loudly. Ranking matches posts to professional interests, so post about the field you're known for (LinkedIn Pressroom).
  • Reply to your commenters, yourself. Real conversation is what LinkedIn says it protects.
  • Invite people you know. LinkedIn's own advice: "only send invitations to people you know and trust," and fewer, more thoughtful ones (Invitation limit reached).
  • Withdraw invitations nobody answers. Pending and ignored invitations count against you (Types of restrictions). Blowhorn can withdraw stale ones for you.
  • Make video match the words. LinkedIn names "videos that don't match the text" as bait.

Don'ts

  • Don't ask for the comment. No "comment YES," no "agree?"
  • Don't join a pod, paid, free, automated or by hand.
  • Don't automate comments. Write each one, and click the button yourself.
  • Don't repost your own post from several profiles at once. That's the pattern LinkedIn describes as a pod. One personal profile per human is the only safe count.
  • Don't recycle last year's hit. LinkedIn is reducing recycled posts.
  • Don't click through a "Take care when connecting" warning. Blowhorn never does (Why flagged invitations are skipped).

Field notes

These come from forums, analysts and our own scars. Nothing here comes from LinkedIn itself.

  • Hypothesis: posts with an outside link in the body reach fewer people than posts without one. Widely repeated by LinkedIn creators; LinkedIn has not said so. Test it on your own posts before you believe it.
  • Hypothesis: the first hour after posting matters most, because early reactions decide how far a post travels. LinkedIn hasn't published this.
  • Hypothesis: invitation limits sit somewhere around 100 a week for most accounts and work on a rolling seven-day window. This is what outreach tools report (one example); LinkedIn publishes no number.

What Blowhorn does on LinkedIn

Blowhorn posts, comments, reposts, invites people to a Page, connects, accepts invitations and withdraws stale ones, from your own Chrome (Platforms). It types at human pace, clicks once, never retries a public action on a guess and hands every CAPTCHA to you (How Blowhorn paces itself). It always skips a flagged invitation. It can still be blocked. Use the dry run first, run LinkedIn slowly, and keep amplify to one personal profile.

More from the Field Manual

  • Reference

    Action matrix

    Every action Blowhorn takes on each platform, and whether that platform's own rules make it safe, a matter of care, or risky.

  • Platform briefing

    X

    What X's For You feed rewards (replies, conversations, time spent, shares) and what sinks a post (mutes, blocks, reports, duplicates, coordinated accounts), straight from the ranking code X published.

  • Platform briefing

    Reddit

    What Reddit rewards (being a real member of a community, early votes, fresh posts) and what gets you removed or banned (spam, self-promotion overload, repeated comments, vote manipulation), from Reddit's own rules and code.

  • Platform briefing

    Hacker News

    What Hacker News asks of posters (curiosity first, your own words, no solicited votes) and why it's the least forgiving front, from its own guidelines.

  • Platform briefing

    Bluesky

    How Bluesky's feeds work (your Following feed, Discover, and custom feeds anyone can build) and the one rule that matters most, no manipulated signals.

  • General education

    Other feeds

    What the other big feeds say they reward and reduce, in their own words. General education only. Blowhorn doesn't post to these platforms.

  • For CLI and agent users

    CLI tuning

    Prescriptive settings for the Blowhorn CLI and blowhorn-axi that map each platform rule in this manual to a flag or default. For operators and the agents they run.