LinkedIn Posting for 90 Days: What I Learned From 270 Posts and 4M Impressions

The Experiment

In Q1 2025, I published on LinkedIn every single day for 90 days. Three posts per day for the first 30 days, then one per day for the final 60. 270 total posts. Nearly 4 million total impressions.

This is what the data showed.

Finding 1: Contrarian Posts Outperform Agreement Posts by 3x

Posts that challenged conventional wisdom consistently got 3–4x the engagement of posts that shared widely accepted advice. Even when the contrarian take was milder, the simple act of saying "here is why the common advice is wrong" triggered more engagement.

Implication: Take positions. Agree less. Push back more.

Finding 2: Hooks With Numbers Outperform Hooks With Questions

Counterintuitively, question hooks ("Have you ever wondered why...?") underperformed compared to number hooks ("73% of consultants do X wrong"). Numbers create pattern interrupts; questions feel like marketing.

Implication: Lead with a specific number whenever possible.

Finding 3: The 8am–9am Window Is Real

Posts published at 8am on weekdays consistently outperformed posts published at other times — but by less than expected (15–20% more impressions, not the 50%+ some claim). The algorithm weight on recency seems real but not dominant.

Implication: Post in the morning window but do not obsess over it.

Finding 4: Comment Volume in Hour 1 Is the Most Important Signal

When I replied to every comment within 30 minutes of publishing, posts received 40–60% more total impressions than equivalent posts where I delayed replying by 3+ hours.

Implication: Clear your schedule for 30 minutes after publishing.

Finding 5: Long Posts Outperform Short Posts for Authority Topics

For posts about strategy, frameworks, and contrarian positions, longer posts (1,200+ characters) significantly outperformed shorter ones. For personal observations and quick insights, shorter posts performed better.

Implication: Match length to content type.