What custom B2B lead data from public records actually is
Most sales teams buy prospect data the same way: rent access to a shared database, export a segment, watch the bounce rates. Custom B2B lead data works differently. Instead of querying someone else's static inventory, you get a collection pipeline built around your definition of a good account — the jurisdictions you sell into, the entity types you care about, the trigger events that mean a prospect is in-market right now.
The raw material is the public record. Companies announce themselves constantly through filings: they register, they license, they permit, they borrow. Each of those events is published by a government office, in a format designed for clerks rather than sales teams. Recordpipe's job is the unglamorous middle: collect those records on cadence, normalize them into one schema, deduplicate them against everything already delivered, and hand your team rows worth calling. The output is an asset you own the shape of, not a subscription that resets when you stop paying.
Trigger-event data beats firmographic lists
A firmographic list tells you a company exists. Trigger-event data tells you when it is about to buy. The distinction matters because timing, not targeting, is where most outbound programs fail — the account was right, the quarter was wrong.
- A new business registration is a company that needs banking, insurance, payroll, software, and suppliers all at once.
- A building permit is capital already committed to a project — the spend is scheduled, not hypothetical.
- A new professional license is a practice opening its doors, with every vendor decision still open.
- A UCC filing names a borrower in motion and the lender who won the deal.
A custom pipeline can carry one of these signals or blend several. Sellers into home services might join permits with new registrations; a commercial lender might watch registrations alongside lien filings. The blend is defined in scoping, against your actual sales motion.
From raw filings to a ranked prospect list
Raw public records are hostile to sales workflows. Names arrive in inconsistent casing, addresses in whatever shape the filing office uses, and the same company appears under several near-identical variants. Delivering that mess to reps is how data projects die.
So the pipeline does the refinement before delivery. Entity names are normalized and matched across filings, so one company is one row. Records are classified — industry, entity type, geography — using the filing text itself. Then the AI analysis layer scores each row against your ideal-customer definition and against your suppression lists, so reps open a ranked queue instead of a raw export. Every row keeps a citation back to its source filing, which means your ops team can audit any record that looks wrong, and your reps can reference the actual event in their outreach. Provenance is the difference between verified prospect data and a guessed database.
Sales intelligence delivered where your team works
A prospect data feed only earns its keep if it lands inside the workflow. Delivery is scoped around your stack: webhook pushes that post new records to any endpoint your team controls, a hosted API your ops engineers can query, or scheduled file drops in CSV or Parquet for teams that ingest on their own schedule.
Cadence is yours too. Daily feeds suit teams selling into timing-sensitive moments like new registrations; weekly rollups suit territory planning. Deliveries can be full snapshots or deltas — only what is new or changed since the last run — which keeps your CRM clean instead of bloated. Contracts are fixed-price, quoted after a $500 scoping that includes a sample built from your real target geography, so you evaluate the actual data before committing.
What a delivery looks like
| Field | Description |
|---|---|
| entity_name | Legal name exactly as it appears on the source filing, plus a normalized variant for matching. |
| entity_type | Corporation, LLC, partnership, or sole proprietorship as classified by the filing jurisdiction. |
| event_type | The trigger event: new registration, license issuance, permit filed, lien recorded. |
| event_date | The date the record published, as issued by the jurisdiction. |
| jurisdiction | State or county where the filing was made. |
| business_address | Address as published in the public filing — no scraped or appended contact data. |
| registered_agent | Agent name and address where the jurisdiction publishes one. |
| industry_class | Category inferred by the analysis layer from filing text and license type. |
| fit_rank | Position against your ideal-customer definition, recomputed each delivery. |
| source_citation | Reference to the public filing this row was built from. |