LinkedIn Algorithm Optimization skill

LinkedIn Algorithm Optimization

Overview

Most LinkedIn underperformance is not a content-quality problem — it's a mechanics problem. Creators post the right message in the wrong format, kill reach with external links and hashtag habits from 2021, and misread a platform-wide reach compression as a personal failure.

This skill applies Richard van der Blom's Algorithm Insights Report — the largest recurring public study of LinkedIn organic performance (the 2026 edition analyzed 1.3M+ posts and 80,000+ profiles) — to diagnose reach problems and design posts that work with the algorithm instead of against it.

When to Use

• "My LinkedIn reach dropped — what changed?"
• "What format should I use for this post?"
• "Best time and frequency to post on LinkedIn"
• "Should I still use hashtags?"
• "Why does my video underperform?"
• "Optimize our company page reach"
• "Build a LinkedIn posting system that the algorithm rewards"

Authoritative Foundations

Richard van der Blom (Just Connecting) — Algorithm Insights Report. Annual data study of LinkedIn organic mechanics, 7th edition (2026). Core structural finding: LinkedIn moved from a relationship graph to an interest graph — follower count no longer predicts reach; topic authority and per-post merit do. Full findings, format rankings, and tactical tables → references/richard-van-der-blom-algorithm.md.

Adam Robinson (RB2B) — proof case for algorithm-aligned founder distribution: daily zero-click posts, hook discipline, and profile-link conversion scaled RB2B to $5M ARR with LinkedIn as the only channel. Playbook → founder-brand → references/adam-robinson-founder-brand.md.

Chris Walker (Refine Labs) — dark social context: most LinkedIn influence is unmeasurable in attribution tools. Pair algorithm optimization with self-reported attribution (references/chris-walker-mental-models.md).

Jessie Lizak (Reveting) — when feed posts are not enough and you need a weekly LinkedIn Live as the content engine (WinsDay model): record conversations, repurpose into clips/carousels/posts, prioritize relationships over tricks. Full runbook → linkedin-live-strategy → references/jessie-lizak-linkedin-live.md. Report default: guest on Lives often; Lizak's playbook for hosting when repurposing is disciplined.

Prerequisites

• Personal LinkedIn profile (and company page if optimizing page reach)
• 90 days of post analytics (impressions, engagement, followers) as baseline
• Defined content pillars (see founder-brand / social-media-strategy)

Step-by-Step Process

Phase 1: Diagnose Before Changing Anything

1. Pull the last 90 days of posts into a sheet: format, length, links, hashtags, tags, posting time, impressions, reactions, comments, saves.
2. Separate platform-wide compression from personal decline. Van der Blom's data showed broad YoY declines (views -50%, engagement -25%, follower growth -59% in the 2025 edition) — benchmark against the trend, not 2023.
3. Flag reach killers: external links in body, 5+ hashtags, 5+ person tags, posting more than once per day, AI-sounding text.

Phase 2: Fix Format Mix

Rank formats by the report's reach/engagement multipliers (full table in references/richard-van-der-blom-algorithm.md):

• Document posts (PDF carousels) — top format: ~1.39x reach, highest engagement rate (~6.6%). Use for frameworks, benchmarks, tutorials.
• Short native video (30-90s) — strong engagement (~5.6%) but average reach dropped below baseline (~0.86x); use for thought leadership, not reach.
• Image + text — the volume workhorse; solid middle performance.
• Text-only — works when sharp and short; weakest default format.
• External links in body — significant reach penalty (roughly 25-60% depending on profile type). Put links in comments or on your profile.

Phase 3: Engineer Post Anatomy

The report's four-phase post system: writing → publishing → nurturing → repurposing.

• Hook: first 2-3 lines decide dwell time. Numbers, credibility, or tension before the fold.
• Length: sharper beats longer in 2026 — 34% of top-performing posts are now under 600 characters (up from 21% in 2024). Dwell time is weighted by completion rate, so cut ruthlessly.
• Keywords over hashtags: posts without hashtags now outperform posts with them by 5-10%. Write topical keywords into the body instead.
• Tags: tag 5+ people and the algorithm treats it as a spam signal if most don't engage. Tag only people who will respond.

Phase 4: Work the Engagement Window

1. Golden hour: reply to every comment in the first 60-90 minutes — comment threads (depth, quality, length) are weighted above reactions.
2. Saves are a ranking signal: design "save-worthy" reference content (checklists, benchmarks, frameworks).
3. Pre/post activity: comment meaningfully on ICP-relevant posts before and after publishing; engagement is reciprocal in the interest graph.
4. Frequency cap: 3-5 posts/week, never more than 1/day.

Phase 5: Company Page and Employee Advocacy

Company page reach grew ~33% in the 2026 data while personal profile reach fell — pages are no longer dead. Run pillar-based page content, engage from the page identity, and pair with employee advocacy (employees resharing with their own commentary, not bare reposts). Amplify proven organic winners with thought leadership ads (see paid-social-strategy).

Output Format

LinkedIn algorithm audit + optimization plan: 90-day baseline table, reach-killer findings, format mix recommendation with multipliers, post anatomy checklist (hook/length/keywords/tags), golden-hour engagement routine, posting calendar (frequency + timing), and a 30-day remeasurement plan.

Quality Check

• Baseline separates platform-wide compression from personal decline
• Every recommendation cites a report finding, not folklore
• Format mix includes document posts for save-worthy content
• No external links in post bodies; links moved to comments/profile
• Hashtags removed or capped; keywords written into body copy
• Golden-hour engagement routine scheduled, not aspirational
• 30-day remeasurement date set with the same metrics

Common Pitfalls

1. Optimizing for followers. The interest graph killed follower-count reach. Topic authority — consistent pillars, keyword repetition — is the new distribution asset.

2. Posting links in the body. The single most common self-inflicted reach penalty. Move links to the first comment or let your profile link convert (Adam Robinson's zero-click pattern).

3. Long-form by default. Dwell time is now combined with completion rate. A 2,000-character post most readers abandon scores worse than a sharp 500-character post they finish.

4. Publishing AI-sounding content. AI-generated text is measurably penalized in both reach and engagement. Use AI for research and editing, not voice.

5. Treating engagement as optional. Ignoring comments in the first hour wastes the algorithm's evaluation window. Nurturing is part of the post, not an afterthought.

6. Chasing every algorithm rumor. The report debunks 20 LinkedIn myths per edition. If a tactic isn't backed by data, treat it as folklore.

Execution Artifacts

• references/framework-notes.md — Framework index and authority routing
• templates/output-template.md — Deliverable shell for agent output
• scripts/check-output.py — Lightweight deliverable validator
• references/richard-van-der-blom-algorithm.md — Report findings, format rankings, tactical tables

Related Skills

• linkedin-live-strategy: Weekly LinkedIn Live shows and repurposing flywheel (Jessie Lizak / Reveting)
• social-selling: Turning LinkedIn visibility into conversations and pipeline
• founder-brand: Founder-led posting system (Adam Robinson playbook)
• social-media-strategy: Platform-wide content formats and sizing
• content-distribution: Repurposing posts across channels