How MonetScope generates opportunity scores
Every opportunity in the catalog is anchored to real public discussions rather than model-invented ideas. This page explains how those discussions become a score.
From source posts to a score
MonetScope generates each opportunity score by ingesting the original source posts and comments from public communities (Reddit, Hacker News, X), then running a multi-stage AI pipeline that extracts structured signals: the explicit pain mentioned, the role of the person mentioning it, the strength of demand language ("would pay for this", "I'd switch tomorrow"), the frequency the same complaint surfaces across independent threads, and the saturation of existing solutions. Those raw extractions are scored on six separate axes — pain severity, willingness to pay, market size, urgency, defensibility, and execution ease — which are then combined into the overall score visible above. Validation score reflects how cleanly the signal survives across independent posts; confidence reflects how much evidence supports the conclusion. None of these scores claim to predict whether a specific founder will succeed — they are pre-discovery signals, not investment recommendations.
What counts as evidence
Every evidence signal attached to an opportunity is a distinct passage from a real discussion — either a top-level post or a high-relevance comment — that the AI pipeline has scored on three dimensions: how directly it confirms the pain, how strongly it signals demand, and how clearly it specifies the role of the person experiencing the problem. Evidence is deduplicated across threads so a single complaint repeated 50 times in one Reddit comment chain does not inflate the count. The full evidence list — verbatim quotes, links back to the original posts, and platform-level engagement metrics — is published on every opportunity page, free and without an account.
How to read a score
Scores are pre-discovery signals, not investment recommendations. A high overall score means the discovery cost of finding a real problem has been reduced — it says nothing about whether a specific team can win the market. Treat the score as a reason to run five customer conversations, not as a verdict.
Browse the opportunity catalog to see the methodology applied to real signals.