LinkedIn Post Optimiser

A data-driven toolkit for scoring, improving and maximising the reach of LinkedIn posts. Built on analysis of 1.8M+ posts and designed for B2B thought leadership.

What's in this repo

linkedin-post-optimiser/
├── research/
│   └── linkedin-algorithm-research.md    # The data behind everything
├── skill/
│   └── linkedin-virality-checker.md      # AI prompt: scores posts 0-100
├── playbook/
│   └── post-accelerator-playbook.md      # Step-by-step publishing guide
├── profiles/
│   ├── strategy-profile-template.md      # Blank template - fill in yours
│   └── example-strategy-profile.md       # Worked example (fictional company)
└── README.md

How it works

The toolkit has three components that work together:

1. Algorithm Research - the evidence base. Data on how LinkedIn's algorithm weights engagement signals, the golden window, dwell time, comment threading and format multipliers. Read this first to understand why the other components work.

2. Virality Checker - an AI prompt that scores draft LinkedIn posts 0-100 across six categories: hook quality, content structure, engagement triggers, content value, technical optimisation and audience calibration. Use it with any AI assistant (Claude, ChatGPT, etc.) by pasting the prompt and your draft post.

3. Post Accelerator Playbook - a step-by-step guide for what to do before, during and after publishing. Covers the 30-minute golden window, cross-engagement strategy and what not to do.

Getting started

Step 1: Set up the Virality Checker in your AI assistant

The checker works with any AI assistant (Claude, ChatGPT, Gemini, etc.). Open a new conversation and paste the following:

I want you to act as my LinkedIn post scorer. Here are your instructions:

[paste the entire contents of skill/linkedin-virality-checker.md here]

The assistant will now understand the scoring framework and how to use it.

Step 2: Create your LinkedIn Strategy Profile

The checker scores posts against six categories. The last one, Audience Calibration, adapts to your specific context using a LinkedIn Strategy Profile - a markdown file that describes your target audience, tone, expertise areas and content themes.

You have two options:

Option A: Build one interactively (recommended)

In the same conversation where you loaded the checker, say:

"I don't have a LinkedIn Strategy Profile yet. Help me create one."

The assistant will walk you through seven sections one at a time, asking 1-2 questions per step. At the end it will present a completed profile for you to save.

Option B: Fill in the template manually

Copy profiles/strategy-profile-template.md, fill in each section and save it. See profiles/example-strategy-profile.md for a worked example of a completed profile.

Step 3: Load your profile (if using Option B or a saved profile)

In your AI conversation, paste your completed profile:

Here is my LinkedIn Strategy Profile. Use this for audience calibration when scoring posts:

[paste the contents of your saved profile here]

Step 4: Score a post

Paste your draft LinkedIn post and ask:

"Score this post."

The assistant will return a score out of 100 with a category breakdown, strengths, improvements, AI tells and anti-pattern flags. Target 70+ before publishing.

Step 5: Follow the playbook

Once your post scores 70+, use playbook/post-accelerator-playbook.md for the publishing workflow. The full playbook is for high-stakes posts; the light version covers everyday publishing.

Tips for ongoing use

• Save your strategy profile as a file so you can paste it into new conversations without rebuilding it each time.
• You don't need to reload the checker every time if your AI assistant supports custom instructions, system prompts or project knowledge. Upload the checker and your profile there for persistent access.
• The research document (research/linkedin-algorithm-research.md) is useful background reading but doesn't need to be loaded into the AI. The checker already incorporates the research findings into its scoring criteria.

Key findings from the research

• The first 30 minutes determine 75% of total reach
• Comments from industry experts carry 5-7x more algorithmic weight
• Posts with 3+ person comment threads get 5.2x amplification
• External links in the post body reduce reach by 25-40%
• Dwell time (61+ seconds) correlates with 15.6% engagement rate vs 1.2% for posts viewed under 3 seconds
• AI-detected content sees ~30% reach reduction

Research sources

• 1.8M+ posts (van der Blom Algorithm Insights 2025)
• 1M posts (Socialinsider)
• 621,833 posts (AuthoredUp)
• 50,000+ posts (LinkIntel)
• 40+ additional sources on LinkedIn algorithm mechanics and B2B engagement

Contributing

If you have additional research data or improvements to the scoring framework, pull requests are welcome.

Licence

MIT