LinkedIn Growth & Engagement Playbook

Hook Science: The First Line Is the Whole Game

The hook is everything before the "see more" fold — ~140 chars on mobile, ~210 on desktop. 60–70% of readers never tap "see more", and LinkedIn's own analytics say 65% decide whether to expand a post based on the opening line alone. Postwise · PostInstantly · GetPostAI

• Rule 4 — the winning pattern: a short, specific, unresolved claim (open loop, no question mark). Curiosity-gap + contrarian hooks hit 6.8%/6.2% engagement — ~2.3× other types.
• Rewrite rule: "What do you think about remote work?" → "Bad managers killed productivity long before remote work did."

• NEVER hooks: "I'm excited to announce...", "In today's fast-paced world...", "I've been thinking a lot about...", "Hope everyone had a great weekend!", "I wanted to share something that resonated..."
• Craft library (adapt, don't copy): the data point ("We tested 3 engagement strategies. One had 300% more impact."); the fast-track ("It took me 3 years to learn this. I'll give it to you in one post."); the how-to with number ("Improve your profile in 5 minutes. Here are the 5 fixes."); mid-scene cold open; "Most DPOs don't know about X" (insider FOMO); bold falsifiable claim; the warning (loss aversion); the confession ("For 2 years I believed X. Here's what I finally figured out."). PostInstantly
• Do's: bold the first line (Unicode bold); write the hook FIRST, then the body; fit the ENTIRE hook in 140 chars; match hook type to goal (story→authority, contrarian→reach+comments, number→saves).

Post Anatomy: Line by Line

Five parts, in order: Hook → Expansion → Body → Payoff → CTA. Pick a framework BEFORE writing. Postory · PostAssembly

• Length: sweet spot 1,200–2,000 chars (~200–350 words) — the dwell-time winner; 150–300 only for razor takes; ceiling 3,000; read-aloud test. Postory — Formatting
• Formatting: bold hook line only; bullets with ONE consistent char; bullets only for 3–6 parallel items; 1–5 emojis, ZERO in the hook; 2–5 hashtags at END (near-zero distribution impact); write in the native composer — Google Docs paste strips line breaks, the #1 killer.
• Frameworks: PAS (Problem–Agitate–Solution), BAB (Before–After–Bridge), Listicle, Story + lesson, Hot take. Pick BEFORE writing.

The Algorithm: 360Brew + Depth Score

LinkedIn replaced its ~30-model feed pipeline with 360Brew, a single ~150B-param LLM (LLaMA-3 base, fine-tuned on LinkedIn data) asking one question: "would THIS reader find THIS post worth engaging with?" LinkedGrow · SocialPilot · Gromming

• Dwell time = PRIMARY signal (Mar 3, 2026): 61+ sec average dwell → 15.6% engagement rate; under 3 sec → 1.2%. Long-form and carousels collect dwell.
• Comment quality v2 (Mar 10): <5 words ≈ zero weight; relevance is AI-scored; thread-generating replies count 3×.
• External links: −45–60% reach in the body; "link in first comment" workaround is partially detected. Keep links rare, first comment only.
• Relevance over recency: posts resurface 2–3 weeks later when a topic matches.

• The 3 gates: quality classifier → first-30-min early engagement → 24–48h Depth Score. 20 comments in 30 min > 50 in 8 hours. Gromming
• Creator Mode: DEPRECATED Mar 17, 2026 — features available to everyone.
• Newsletters: boosted March 2026 (push, 48h window, search placement).

The AI-Slop Penalty (July 2026 — THE Rule)

July 30, 2026: LinkedIn ships the "Seems like AI slop" report button, removes the "enhance your post" AI writer (proofread-only now) and ramps classifiers. TechCrunch · Fortune · LinkedIn Newsroom

• Definition: "polished but lacks substance — no experience, perspective, or insight." Using AI to refine language is FINE. "AI and slop are not the same thing" (CPO Srinivasan).
• The M~ rule: Zeus drafts → M~ rewrites in his own voice. AI = first-draft accelerator, never the publish button. LinkedIn Newsroom

The Psychology of Engagement

The algorithm is a mirror of human psychology: every notification is a dopamine hit, comments reward the author, and reciprocity compounds. CommentRocket 1 · CommentRocket 2

• Dopamine loop: every notification = a hit; comments reward the author → reciprocity works.
• Why people COMMENT (effort matters): (1) STATUS — trigger a question that lets them show expertise ("DPOs: the one AI tool you'd ban?"); (2) IDENTITY — value statements ("I never ship AI features without a DPIA" → "Agree!"); (3) ALTRUISM — be vulnerable, ask for help.

• Mere Exposure Effect: repeated valuable appearances = "safe and trustworthy"; familiarity breeds trust, trust breeds follows. Consistency beats intensity.
• Specificity detector: "Many professionals struggle with X" = invisible. "You've rewritten that email four times and it's not even noon" = stops the scroll. People engage when they feel SEEN.
• Saves = highest intent: give frameworks, checklists, ladders, RACI structures — SAVE-bait.
• Niche finding (Linqin, 16,000 comments): posts/comments referencing governance, guardrails, data lineage, ownership/RACI score higher credibility and drive discussion; repeatable frameworks beat hype; real KPIs beat generic metrics; authenticity + transparency = highest trust fuel. Linqin Study

Commenting Strategy: The Compounding Move

360Brew reads your comments like posts — every substantive comment is part of your topical signature. Co.Actor · FeedsSquad

• Substantive = personal experience, counter-perspective, extended framework, substantive question, specific example. Sweet spot: 2–5 sentences, <40 words (17.1% reply rate). Test: "would a stranger benefit from reading this?"
• NOISE (penalizes YOUR profile): "Great post!", "100%!", tags without context, self-promotion, "Agree?", AI templates, pod comments, rings.
• TIMING: first 1–3 hours = 23.2% reply rate vs 16.2% baseline; first hour = 4–10× visibility multiplier; early = top-of-thread + author online; first-hour beats next-day by ~113%. Linqin
• Routine: 15–20 min/day, 10–15 substantive comments. Two-layer target list: (1) your ICP in-niche, (2) niche creators 1k–50k followers (NOT mega-influencers — a comment on a 1,500-view post in your category outperforms the same on a 50,000-view post).
• 90-day topical consistency or the signal dilutes — comment WITHIN your lane.

Timing, Cadence and the Golden Hour

The first 60–90 minutes are a 2–5% sample of your network; velocity + dwell in that window lock the distribution tier for 48h. Only ~5% of underperforming first-hour posts ever recover. LinkBoost · LinkedIn Insider · DistributionBase

• YOUR golden hour = ICP online × you available to reply. Test 2–3 windows for 2 weeks, lock the winner. (SG-based serving EU/US: Tue/Thu ~10am SGT or 4–5pm SGT.)
• Cadence: 2–5×/week (3× target: 2 text + 1 mascot/tip). NEVER two posts within 4 hours. Consistency > virality.
• Dwell economics: 61+ sec → 15.6% ER; sub-3 sec → 1.2%. Long-form and carousels (8–12 slides, 55s dwell) collect dwell. Buffer

Formats: Carousels, Video, Text

Format choice is a distribution decision — carousels are the engagement king, video is Phase 2, text stays the cadence backbone. Gromming · Chime

• Video (Veo 3): commercial OK; Google AI Pro = 10 free generations with a visible watermark (fine for tests, not brand); Ultra/API = clean. Tag AI video (C2PA / EU AI Act Art 50). Video hooks = same rules as text — on-screen claim within 2s. Video = Phase 2, never blocks the text cadence.

The AI Governance Niche Playbook

The lane is perfect for 2026: scarce, high-stakes, decision-maker audience; governance/guardrail content scores the highest credibility + discussion (Linqin); saves + authority + status all stack. Linqin

• Conversion levers: status questions for DPOs; identity value statements; saves via ladders/checklists; data authority ("We audited 30 SMEs' AI inventory"); vulnerability; narrow questions to lurkers.

Trending Topics: August 2026 (verified, citable)

Six verified stories dominate the governance feed this month — each with a ready hook angle. AI CERTs

Weekly Loop, 12-Week Plan, KPI Benchmarks

A weekly operating loop, a 12-week build plan, and the only KPIs that matter — tuned to the 2026 algorithm. Buffer · TryOrdinal

• Weekly loop: Mon source-watch → 3 topics → 3 drafts (Zeus drafts, M~ rewrites, slop audit); Tue/Thu/Fri golden-hour posts; DAILY 15–20 min commenting; Sun 30-min numbers review (impressions, saves, dwell, comments, followers, per-format).

All Sources (37)

Every claim in this playbook traces to one of these. Open the relevant chapter for inline citations.