Prospect Enrichment & Outreach Review

Goal

Clean, enrich, and validate raw B2B prospect/company lists, apply ICP or niche filters, create sourced personalization notes and send-ready outreach drafts, triage prior replies, and deliver structured files plus a human-approval checklist.

Inputs to gather

• Raw prospect/company files (URLs/domains, names, titles, emails) — source data to enrich
• Email-finder results (email addresses + validity status) — needed to confirm deliverability
• ICP/niche filters (industries, size, regions, personas, exclusions, compliance rules) — drives keep/reject decisions
• Outreach parameters (product/offer, value props, template variants, tone, CTA, mandatory/legal text) — governs draft emails
• Prior campaign reply logs (threads or CSV, with timestamps and owners) — prevents re-contact and sets follow-ups
• Any preferred output column mapping or file naming conventions — ensures compatibility with downstream systems

Before doing any work, ask the user for these inputs in ONE message. Skip anything they already provided. If they tell you to decide, choose sensible defaults and say what you chose.

Workflow

1. Confirm scope
   a. Restate ICP/niche filters and outreach parameters to the user for approval.
   b. List received files; note any missing items and request them if critical.

2. Load & normalize (Read)
   a. Ingest all provided files; standardize columns and casing.
   b. Normalize domains (lowercase, strip protocols, resolve obvious redirects).
   c. Unify into one working table; de-duplicate by email, then name+domain, then domain.

3. Fill missing domains
   a. When company name lacks a website, use WebSearch to locate the corporate site; confirm via branding in snippets.
   b. Store alternates but choose one canonical domain per company.

4. Company enrichment (WebFetch/WebSearch)
   a. Fetch homepage + About/Careers/Products pages; extract industry, positioning, HQ, markets, tech stack, recent launches, partnerships, compliance claims.
   b. Validate headcount and funding stage from open sources only; mark Unknown if unsure.

5. Contact validation & enrichment
   a. Merge email-finder results, attaching email, email_status, and source.
   b. Normalize titles to role, seniority, department; flag mismatches with target personas.
   c. Verify geography when required.

6. Apply ICP/niche filters
   a. Encode rules from the user’s criteria.
   b. Tag each record icp_fit = Good, Borderline, or Out with icp_reason.
   c. Reject any hard fails (invalid email, out-of-ICP, prior unsubscribe, bounce, consumer-only domain).

7. Build personalization notes
   a. Extract 1–3 verifiable, timely signals per company/contact.
   b. Write a 1–2-sentence personalization_note citing the signal and its relevance; include source URL.
   c. If no credible signal: “No recent credible signal found.”

8. Draft outreach
   a. Select outreach_template variant based on persona + signal.
   b. Create a 3–6-word subject_line and an 80–120-word email: greet, reference signal, state value prop, single CTA, include compliance copy.
   c. Use merge tags like {{first_name}}, {{company}}, {{signal_snippet}}; ensure claims are factual.

9. Review prior replies
   a. Map reply logs to current records.
   b. Categorize replies (Interested, Referral, Not Now, OOO, Unsubscribe, Bounce, Existing Customer, Not a Fit, Competitor).
   c. Set reply_route, next_action, owner, and due_sla per playbook; label do_not_contact where applicable.

10. Quality assurance
    a. Validate mandatory fields for Good/Borderline records.
    b. Spot-check 10 random records for signal accuracy and draft truthfulness.
    c. Ensure merge tags resolve and no duplicates will be emailed.
    d. Summarize counts: total, enriched, Good, Borderline, Out, top rejection reasons, email validity %, reply categories.

11. Produce outputs (Edit)
    a. enriched_prospects.csv with all enriched and classified fields.
    b. rejects.csv listing disqualified records with reasons.
    c. personalization_notes.csv/md containing note and source URL per company/domain.
    d. outreach_drafts.csv with subject_line, draft_email, and template_variant.
    e. reply_routing.csv with reply_category, owner, next_action, due_sla, do_not_contact.
    f. approval_checklist.md featuring the Human Approval Checklist and summary metrics.

12. Check in with the user
    Provide the summary counts and the approval_checklist.md for review before final hand-off of all CSV/MD files.

Output

A zipped package (or shareable folder) containing:

1. enriched_prospects.csv
2. rejects.csv
3. personalization_notes.csv or .md
4. outreach_drafts.csv
5. reply_routing.csv
6. approval_checklist.md (includes the human approval checklist and summary metrics)

Files must be UTF-8, comma-separated (or Markdown where noted), and ready for immediate import into CRM or sequencing tools.