Source Confidence
Source confidence is a score or label indicating how trustworthy a specific data point is based on the quality and strength of its evidence.
Allocator relevance: Prevents users from treating weak signals as facts—critical for mandate fit, ownership, and decision-maker claims.
Expanded Definition
Not all sources are equal. Source confidence reflects whether a field is confirmed by primary evidence (official filings, direct confirmation, authoritative disclosures) versus inferred from indirect signals (news mentions, social profiles, third-party lists). High-quality systems separate “confirmed” from “inferred,” attach lineage, and show last verified.
For allocator intelligence, confidence scoring reduces reputational risk (mislabeling principals, misrouting outreach, misrepresenting mandates).
How It Works in Practice
Platforms assign confidence tiers per field based on evidence type, recency, corroboration, and consistency. Users can filter by confidence and decide where human verification is required.
Decision Authority and Governance
Governance defines confidence rules, evidence requirements, and escalation paths for disputed data. Without consistent rules, confidence labels become marketing.
Common Misconceptions
- Confidence means “true.”
- A single source is enough if it’s recent.
- Confidence scoring can be applied uniformly across all fields.
Key Takeaways
- Confidence is evidence-weighted, not opinion.
- Field-level confidence is more useful than record-level.
- Pair confidence with lineage and recency.