SEC EDGAR full-text search: what it returns, and whether it works in Clay

SEC EDGAR full-text search is not available in Clay. It is best for Buying signals, Diligence. Pricing: Free. Access: Free JSON endpoint behind the UI, plus UI. SEC fair-access limit of 10 requests per second with a declared User-Agent. The endpoint is not a formally documented public API..

Data domains: Corporate events, Company, Signals, Funding and finance.

What does SEC EDGAR full-text search return?

Official fields: (hits._source) ciks; display_names; form; root_forms; file_date; period_ending; adsh (accession number); file_num; film_num; file_type; file_description; sequence; biz_locations; biz_states; inc_states; sics; items (8-K item numbers); xsl.

What can you derive from it?

1. Pain disclosure: filings mentioning a problem the portco solves, such as material weakness, ERP implementation or a cybersecurity incident. 2. Event timing: 8-K items and file_date give the exact date a trigger became public. 3. Competitor mention: filings naming a competitor product as a vendor or risk. 4. Industry cluster: sics and biz_states group matches into a segment. 5. Repeat mention: the same phrase across successive 10-Ks shows a persistent issue.

How would a GTM team use it?

1. Revenue trigger: alert on new 8-Ks or 10-Ks containing a pain phrase and route them to outreach within days of filing. 2. Competitive watch: track filings that name a competitor's product to find its public customers. 3. Buyer map: pull signers and officers from matched filings for executive outreach. 4. PE screen: search S-1s and 10-Ks for mentions of a portco category to find add-on targets and their customers.

What are the caveats?

Not a native Clay integration as of 2026-10-17 (checked against Clay's integrations catalog). Can be called through Clay's HTTP API column where an API exists. External observation: filing language shows disclosure, not budget or buying intent. Covers filings from 2001 onward and only SEC filers, so most private companies are absent. The JSON endpoint is undocumented and can change. Phrase matches need human review because boilerplate risk language produces false positives.