When to use websearch enrichment

Question: "Find phone and address for Joe's Plumbing in Tampa."

Two paths

Situation Tool
You have a phone number enrich_phone ($0.330) or enrich_person with phone ($0.500)
You have name + location enrich_person ($0.500)
You only have business name + city/state enrich_websearch ($1.00 min)

enrich_company and enrich_email are catalog-tier and offline by default.

Tool

POST /api/v1/enrich/websearch or MCP enrich_websearch

{
  "business_name": "Joe's Plumbing",
  "city": "Tampa",
  "state": "FL"
}

Optional domain when known. Omit prompt to use the benchmark hunt_v2 phone-hunt instructions (review-aggregator sites, owner cells, source URLs). Pass prompt only when you need custom research instructions.

Workflow

  1. Try the cheapest narrow tool first (phone → enrich_phone; person name → enrich_person).
  2. Escalate to websearch when structured lookup lacks coverage for a local business name.
  3. Report structured fields in data — phones, people (name+title), addresses, and legal_associates when present.
  4. If matched: false, do not fill gaps from general web knowledge.

data.legal_associates flags individuals legally tied to the business — attorneys of record, registered agents, officers on state filings, family counsel, or family members acting in a legal capacity. Lead users need this to avoid compliance issues (cold-calling attorney cells, mistaking registered agent for owner, TCPA/consent documentation).

Each entry:

{
  "name": "Richard Shenkan",
  "role": "family_attorney",
  "relationship": "Family member; PA-licensed attorney",
  "phone": "+14127165800",
  "source": "https://...",
  "outreach_note": "Attorney associated with ownership family — prefer storefront line before direct cell"
}

Agent behavior: when legal_associates is non-empty, surface it prominently in the lead summary — not buried under generic people. Distinguish operational owners (people) from legal/counsel contacts (legal_associates).

Cost note

Websearch is priced higher ($1.00 minimum on match; 30% markup when COGS > $1) because it runs the managed research stack (search + synthesis). Use it deliberately, not as a default for every lookup.