The LinkedIn Algorithm Update 2026: 360Brew, Depth Score, and the End of the Engagement Economy

If you landed here searching for a '2025 LinkedIn algorithm update,' the honest answer is that the document you were looking for is already obsolete. Between June 2025 and April 2026, LinkedIn didn't tweak its ranking system — it replaced it. Velocity-based ranking is gone. Engagement pods are detected with 97% accuracy. The first hour still matters, but for a very different reason. This is the full timeline, what each change actually does, and the playbook that survives the new rules.

The 2025–2026 algorithm timeline: what shipped, when

The rollout was not a single switch. It was a nine-month sequence of quiet deployments, each of which invalidated a previous best-practice deck.

  • January 2025 — LinkedIn Engineering publishes the 360Brew paper on arXiv, describing a 150-billion-parameter decoder-only foundation model capable of performing 30+ ranking and recommendation tasks. Most marketers ignored it. They shouldn’t have.
  • Spring–Summer 2025 — 360Brew is tested on a slice of feed traffic. Our original version of this article (June 2025) was written against the previous ranker; the ‘reach down 50%, engagement down 75%’ stats from Richard van der Blom’s May 2025 report reflect that legacy system’s decline.
  • September–November 2025 — LinkedIn issues public statements that it will take further action against engagement pods. Lempod is pulled from the Chrome Web Store. Accounts using automation see reach drops from 8,500 to 340 impressions overnight.
  • December 2025 – January 2026 — 360Brew handles the majority of feed scoring. Forbes journalist Jodie Cook visits LinkedIn’s NYC headquarters and publishes her 12 January 2026 investigation, the first major-publication confirmation that the old ranker is effectively retired.
  • March 2026 — the ‘Authenticity Update’. This is the one that hurts. LinkedIn ships a detection model for pod-like behaviour with 97% accuracy, kills external-link-spam boosts, and makes the Depth Score — a composite signal that measures silent reading behaviour — the primary driver of reach.

If you are still optimising for the 2025 rules, you are competing in a sport that ended.

What 360Brew actually scores (and why it broke old playbooks)

360Brew is not a bigger version of the old EdgeRank-style scoring system. It’s a foundation model that evaluates your post the way a well-briefed human would: semantically.

The practical consequences:

  • Keyword stuffing does nothing. 360Brew reads meaning, not strings. Cramming ‘AI, AI agents, AI marketing, AI automation’ into the first 70 characters used to be a cheap win. Now it reads as spam.
  • Generic comments are worth zero. The model can tell the difference between ‘Great post!’ and a reply that references something specific you wrote. The second gets credit; the first doesn’t — and a pile of the first actively flags your post as suspicious.
  • Niche relevance compounds. 360Brew tracks what each viewer consistently engages with deeply, not just what they click. If your post is genuinely about B2B pricing and the reader’s Depth Score history says they read pricing content, you get surfaced. Broad ‘thought-leader’ slop gets buried because it matches no niche profile.
  • Cross-posting tricks are dead. Duplicating personal posts to company pages within 24 hours (the old 2025 hack) is now pattern-matched as low-effort syndication.

This is why your 2025 playbook stopped working around February. The signals didn’t weaken. They were replaced.

The Depth Score — LinkedIn’s new primary signal

The Depth Score is the headline change in the March 2026 Authenticity Update. Instead of counting reactions, the algorithm measures silent attention:

  • How long a viewer pauses on your post without scrolling
  • Whether they tap ‘see more’ to expand the body
  • How long they spend in the comments (and which comments they read)
  • Whether they save the post
  • Whether they share it privately via DM (weighted higher than public reshares)

Public likes and generic comments — the two things engagement pods manufacture — are now near-worthless. This is intentional. We covered the enforcement mechanics in detail in our cousin post on the LinkedIn engagement pods crackdown and the engagement-surge detection system that flags pod spam.

If your content is skim-friendly but not worth reading, your Depth Score is low and your reach collapses. If your content makes people stop scrolling for 14 seconds, you win — even with fewer reactions.

The 60-minute window still decides lifetime reach

One thing survived from the 2025 playbook: the early window matters.

But it’s been redefined. In 2025 it was the ‘Golden Hour’ of likes-and-comments velocity. In 2026, the first 60 to 90 minutes determine up to 70% of a post’s lifetime reach, and the signal the algorithm looks for is substantive, multi-sentence comments from accounts that have no prior reciprocity pattern with you. If those land in the first hour, 360Brew’s Momentum Model flags the post as trending-within-niche and expands distribution geometrically through hour eight.

If the first-hour comments are five-word emoji replies from the same fifteen accounts that comment on every one of your posts — that’s a 97%-accuracy pod flag, not a virality signal.

Tactical playbook: how to post on LinkedIn in 2026

This is the section we rebuilt from the studs. The original June 2025 advice — 300-character posts, vertical video, alt text on slides — still has partial validity, but the weighting has changed.

  1. Open with a specific claim, not a hook-bait line. ‘Here’s a controversial take’ is 2024. In 2026, 360Brew parses your opening semantically — lead with the most specific sentence of the entire post and put a concrete number in it where possible.
  2. Write for the Depth Score, not the like button. Make the post worth reading twice. That usually means one counterintuitive claim, two pieces of evidence, and a question that requires actual thought to answer.
  3. Post when your niche audience is awake — not when ‘best time’ dashboards say. 360Brew’s distribution is niche-gated; volume-based timing advice no longer applies.
  4. Reply to every first-hour comment with a 2+ sentence reply that continues the argument. This is the single highest-leverage activity in 2026. It extends dwell time on the comment thread, which is a primary Depth Score input.
  5. Kill the pod. If you’re still in one, leave today. We explain the detection mechanics and the 60–90 day recovery window in our companion piece on LinkedIn’s engagement-surge crackdown.
  6. Short doesn’t mean thin. The micro-content strategy we published as an alternative to long-form still works — but only if each short post carries one hard-earned insight. Empty short posts score worse in 2026 than they did in 2025.
  7. Vertical video still wins — if it’s under 90 seconds and captioned. 360Brew can now transcribe and score video semantics. ‘Style over substance’ edits actually hurt you.
  8. Forget cross-posting hacks. Publish company-page and personal-profile content with distinct framing, 48+ hours apart, or don’t bother with the company page.

What this means for marketers

The strategic picture, stripped of jargon: LinkedIn has stopped rewarding activity and started rewarding attention. You cannot manufacture attention with a rota of fifteen friends anymore. The only durable lever is content that earns the 14-second pause.

For in-house marketers and agencies, this raises an uncomfortable question. The old ‘LinkedIn growth’ service — pods, scheduled reposts, cross-promotion rings, generic ‘engagement comments’ — is now a liability. Anyone selling you that in April 2026 is selling you a shadow-ban.

The work that remains — genuine point-of-view development, fast response to first-hour comments, multi-platform production of the same insight in different formats — is the work that has always been expensive. It’s also the work AI is now genuinely good at doing at scale, if you give it the right operating system. That is where Anjin fits.

Anjin: the Marketing Operating System built for algorithm-proof output

Anjin is the Marketing Operating System — a single workspace where AI agents draft, format, schedule and respond to content across every channel your audience actually uses. It isn’t a LinkedIn automation tool. It doesn’t fake engagement. It doesn’t run pods.

What it does, specifically for a 2026 LinkedIn reality:

  • Drafts posts with one specific claim plus supporting evidence — the exact structure 360Brew rewards — from your raw notes, calls, or source material.
  • Generates five variants of each post for different audience slices, so niche distribution actually has something to bite on.
  • Monitors first-hour comments and drafts substantive, context-aware replies you approve in a single click — compressing the Depth Score window from a two-hour chore to a ten-minute pass.
  • Repurposes the same insight into short video scripts, carousel copy, and a blog follow-up so you stop treating LinkedIn as its own silo.

One operator plus Anjin replaces the four- or five-person content team whose job was to feed the old algorithm. You are now feeding a different one. The workflow has to change.

The £888 Lifetime License — Offer Closing Soon

Lifetime access to Anjin for a one-time payment of £888. Not a subscription. Not a seat. Not a trial. One payment, unlimited use, for as long as Anjin exists.

The average marketing team spends £888 in about three working days on tooling, freelancers and coordination software. You’re buying the platform that replaces most of it — once.

This price will not be offered again once we close our early-access cohort.

Claim your £888 Anjin lifetime license →

Founders, agency owners and in-house marketers — this is how you run marketing at AI speed without the team, the burn, or another year of waiting.

Sources: arXiv 2501.16450 (360Brew), Social Media Today (LinkedIn pod enforcement), LinkBoost (2026 algorithm changes), UpGrowth (360Brew explainer), Growleads (60-minute window), ConnectSafely (pods crackdown)

Continue reading