A signal is not a fact until the evidence supports it.

New Practice Intel is designed to reduce false positives by separating raw record changes from commercially meaningful events.

1. Detect candidate events

Candidate records may originate from sources such as CMS provider data, state business registrations, professional licensing databases, public permits, business websites, business listings, hiring activity, and opening announcements.

2. Resolve entities

Names, addresses, organizations, providers, and related records are normalized so that routine duplicates or pre-existing entities are not mistaken for new businesses.

3. Corroborate independently

High-confidence leads should generally require at least two independent source signals. Stronger evidence may include a new organization record plus a new business entity, commercial address, licensing action, permit, website, or other corroborating event.

4. Score confidence and stage

Each accepted lead is assigned an explainable confidence score and likely business stage. Contradictory evidence, stale records, residential addresses, and indicators of established operations reduce confidence or cause rejection.

5. Preserve source context

Where practical, records retain source URLs or source identifiers so customers can understand why a signal was surfaced.

Important limitation

Public records can lag reality or reflect administrative changes rather than real-world business events. New Practice Intel is decision-support intelligence, not a guarantee that a practice is new or that a purchase will occur.