Add a LinkedIn Post Performance Analyser with niche benchmarking and AI optimisation

Add a Post Performance Analyser to the LinkedIn content workflow that evaluates a post before publishing and compares it against relevant high-performing posts from the same niche.

The goal is not to show a generic AI-generated “virality score”. The analyser should use real benchmark signals wherever data is available and clearly distinguish measured data from AI predictions.

The analyser should include:

• Predicted Performance Score (0–100) based on the post, historical account performance where available, and niche benchmarks.

• Estimated Reach shown as a range rather than false precision.

• Estimated Engagement Rate with the assumptions behind the estimate.

• Niche Benchmark Comparison against relevant high-performing LinkedIn posts from the same topic/niche.

• Show 3–5 benchmark posts with useful metrics such as reactions, comments, reposts, impressions/reach where available, author/account size where available, and publication date.

• Explain why each benchmark is comparable to the current post rather than simply showing popular posts.

• Analyse the hook, structure, readability, post length, formatting, topic relevance, CTA, media usage, and engagement patterns.

• Provide specific recommendations ranked by likely impact.

• Generate an Optimised Rewrite while preserving the author’s original meaning, facts, voice, and positioning.

• Allow users to compare Original vs Optimised before accepting changes.

• Add Apply & Preview Rewrite so the user can send the improved version back into the LinkedIn post editor without overwriting the original until confirmed.

Important: avoid fabricated performance claims. If LinkedIn or another connected source does not provide enough data to support a reach, engagement, or benchmark prediction, label it clearly as an AI estimate and show confidence/available evidence rather than presenting it as measured fact.

The analyser should become part of the existing LinkedIn creation workflow, not a disconnected standalone tool.