How to Write a LinkedIn Hook That Doesn't Sound Like AI

In 2023, AI-generated LinkedIn posts looked novel. In 2026, they're a signature the algorithm and the reader both recognize.

The AI tells

If your post opens with any of these, you're flagged — both by the algorithm's freshness model and by readers:

• "In today's rapidly evolving landscape..."
• "As we navigate the complexities of..."
• "Here are 5 key takeaways..."
• "Unlocking the potential of..."
• "🚀 Excited to share..."
• "Game-changer."
• "Thoughts?"

Seven hooks that still work

1. The concrete number

"We shipped 14 features last quarter. Two drove 80% of new revenue. Here's what those two had in common."

2. The contrarian claim

"Most advice on scaling SDR teams is wrong. Not because it's fake — because it was written for 2018."

3. The mistake confession

"I made a $40k mistake yesterday. Here's the decision tree I should have used."

4. The specific scene

"Last Tuesday, 4pm. A customer called asking for a refund. What I said next 2x'd their contract."

5. The reframe

"We don't have a hiring problem. We have a firing problem — we wait 6 months too long."

6. The pattern observation

"Every B2B founder I've met who crossed $10M ARR did this one uncomfortable thing."

7. The forbidden take

"Unpopular opinion: most 'data-driven decisions' are post-hoc rationalizations."

What they have in common

Every hook above has:

• A specific number, scene, or person in the first sentence
• Tension — a problem, mistake, or contradiction
• No corporate vocabulary — no "leverage," "unlock," "journey," "landscape"

Write it, then cut the first sentence

The fastest way to fix an AI-sounding hook: write the whole post, then delete the first sentence. Most AI drafts use the first sentence to warm up. You don't need to warm up on LinkedIn. You need to start.

The rewriting heuristic

If a stranger read only your first line, would they feel like they have to read line 2? If not, cut it.