Eduard Lakhno · 8 minLet’s speedrun outbound in 31 steps.
Enough to build the process from scratch for your own product or service.
1. Focus on 1, max 2 ICPs. Wider and hypotheses stop being testable.
2. Hypothesis = offer (what) + ICP (who) + angle (how).
3. Change no more than one variable at a time, or the result tells you nothing.
4. Budget ~2,000 unique contacts per hypothesis across 500+ companies.
5. Build the list company-first: higher accuracy plus account-level signals.
6. LinkedIn: higher conversion, harder to scale; ~500 connects/mo per sender.
7. Cold email: lower conversion, deliverability hassle, but faster to validate; ~350 emails/mo per inbox.
8. Email hygiene starts with a separate .com domain and Google inboxes on it.
9. Inboxes straight from Google Workspace (pricier, safer) or resold via Premium Inboxes/Zapmail (cheaper), max 3 per domain.
10. Warmup runs permanently, 50/50 with campaigns, sending starts week 4.
11. LinkedIn hygiene: 20-30 connects a day, SSI 50+, acceptance rate above 15%.
12. Spin up a separate database on Supabase ($0/m) before any CRM and push campaign logs there via webhooks.
13. A simple CRM lives there too: views and reports in the Supabase UI, Claude builds the base over MCP and reads it back.
14. Then two semi-technical problems: where to get the data and how to send.
15. Data comes in two layers: companies and the people inside them.
16. Companies: Apollo Basic ($49/m), enrichment via n8n ($9/m) + Supabase ($0/m).
17. People are the starred task: you need current employer, title and working email.
18. The single largest source is LinkedIn, and it caps the volume.
19. Coresignal, Brightdata and Harvest solve that with a proxy endpoint: send a domain and titles, get people back.
20. Per 1,000 contacts with email: Coresignal $34, Brightdata $11.6, Harvest $10.
21. Out comes two linked lists: companies and people.
22. Run the list through several campaign variants to find what works.
23. Write the draft by hand, leave AI for light editing per recipient.
24. No secret frameworks: start with {{sender_intro}} + {{trigger}} + {{consequence}} + {{offer}} + {{micro_ask/CTA}}.
25. Filled in:
{{sender_intro}}: hi, my name is Eduard
{{trigger}}: I see you read my post this far
{{consequence}}: so you want better outbound results
{{offer}}: happy to audit your process
{{micro_ask/CTA}}: message me, I’ll share examples
26. The key is {{trigger}}: it decides everything that follows, and whoever learns to find good ones has outbound solved.
27. Sending tools: Smartlead for email, Sprootal for LinkedIn.
28. Log key events: connection sent, connection accepted, message sent, reply received; email sent, reply received, bounce.
29. That log is your analytics base and keeps progress in one place.
30. PRR = positive replies / contacted, how the hypothesis performed.
31. CPL = outbound costs / leads, the economics of outbound.
Outreach data is the core asset, and from there it feeds your CRM.
What would you add? Drop it in the comments.
