CI-301b · Module 2
Social Signal Collection
3 min read
Social platforms produce the highest volume and lowest signal-to-noise ratio of any OSINT category. The value is in the unfiltered perspective — employees discussing company culture on Glassdoor, engineers debating technology choices on Hacker News, executives signaling strategic intent on LinkedIn. Collection infrastructure must handle the volume while filtering for the signal. Keyword-based collection with relevance scoring is the baseline. AI-assisted sentiment and intent classification is the production standard.
Do This
- Monitor executive LinkedIn activity — posts, comments, and connection changes signal strategic intent
- Track Glassdoor review trends, not individual reviews — the trajectory reveals more than any single opinion
- Use keyword tracking on Hacker News and Reddit for technology adoption and competitive sentiment
- Score social signals for relevance before they enter the analysis pipeline — most are noise
Avoid This
- Treat individual social media posts as intelligence without corroboration — one person's opinion is not a finding
- Ignore social signals because they are noisy — the noise contains signal worth extracting
- Monitor everything without filtering — the volume will overwhelm your analysis capacity
- Use social signals for quantitative claims — they are for sentiment and intent, not for numbers