Cluster LinkedIn post analytics by content type and collateral type, not just media format

Problem

Today post performance is grouped by media format only (image, carousel, video, text). That does not match how creators actually think about their content. I categorise my posts mentally as thought leadership, rants, personal stories, event-related (SaaSiest, certifications, meetups), and so on. I already know from gut feel that event posts with a group photo land at roughly X impressions, personal stories do better, and rants do best on both impressions and reactions. But I have no way to see this in the data, so I cannot confirm it, size the gaps, or plan content around it.

What I want

Two independent tags on every post, with performance clustered along both axes:

1. Content type (what the post is about): thought leadership, rant, personal story, event-related, product/promo, educational, other. Custom tags should be possible.
2. Collateral type (what the visual is): selfie, photo with people, infographic, screenshot, native video, carousel, text-only, other.

Then a view that shows average impressions, reactions, comments and engagement rate per content type, per collateral type, and for the combination (for example "rant + selfie" vs "rant + text-only"). Those combinations are where the real insight is: it is content times collateral that drives performance, not either one alone.

Nice to have

• Auto-suggest the content type and collateral type for past and new posts, with a one-click correction so the user keeps control over the labels.
• Trend over time per cluster, to see whether a format is wearing out.
• A "what to post next" hint based on which clusters are underused relative to their performance.

Why it matters

It turns LinkedIn posting from intuition into a repeatable content strategy, and it gives creators a concrete reason to keep their analytics inside SurroundR instead of a spreadsheet.