← Files AIsa GTMARCHIVED FILE
skills/prospecting/README.md
2.49 KB · Oct 5, 2026 · 18:25 UTC
# Prospecting Prospecting turns a defined B2B ideal customer profile into a bounded, ranked list of companies and, when requested, relevant business decision-makers. Every recommendation carries the available evidence, source date, confidence and unresolved data gaps. ## What it can deliver - a pass/fail ICP checklist and explicit ranking method; - net-new account discovery or qualification of an existing company list; - company-domain deduplication and finalist-only enrichment; - decision-maker role research within shortlisted companies; - evidence-backed fit, timing, contactability and confidence; - a concise table of up to 25 accounts by default, or a bounded file-ready result when requested; - costs, partial failures, unknowns and verification steps. ## Example requests > Find 15 US B2B SaaS companies with 100–500 employees that use Salesforce, explain each match, and show the research quote before any paid execution. > Qualify this existing domain list against our ICP. Reuse the fields I supplied and enrich only finalists with material gaps. > For these five approved accounts, identify likely revenue-operations decision-makers. Do not reveal private emails or phone numbers. > Use these two best customers as lookalike seeds, then re-check every candidate against our ICP instead of treating similarity as fit. ## Data and execution boundaries The first release focuses on ordinary B2B and B2B SaaS. Apollo is the principal company/person source; it is not a multi-provider waterfall. Public website research and Similarweb signals are conditional, not default checks for every account. The skill searches a small sample, shortlists before enrichment, and prices exact AIsa calls before asking for execution approval. Every paid page, enrichment, contact match or conditional signal requires a matching quote and explicit authorization. It never retries paid calls or paginates automatically. Search results, enrichment, `has_email`, and returned addresses are not verified deliverability or buying intent. Missing results remain unknown. Company fit and a person's role fit are scored separately, and sensitive attributes are never inferred or used. This skill produces research only. It does not send email, enroll contacts in sequences, change CRM or Apollo workspace records, monitor signals continuously, or create battle cards. Customer-need research belongs to Customer Research, competitor comparison to Competitor Profiling, outreach writing to Cold Email, and sales collateral to Sales Enablement.
SHA-256: a85f9e72f822a0710a6e0c5bcc365d867254a9700c98fbb003c0046ec3b11051