Raw data from 6 independent sources was collected, cross-referenced, and enriched through a 7-stage AI pipeline — converging into a single per-contact intelligence record covering each person's company, role, tech context, and personalised sales angle.
Infrastructure wired for email and LinkedIn outreach — sending accounts, domain warmup, campaign tooling, and CRM sync.
Multi-sender campaign management. Contacts enrolled across sequences segmented by buying role, tier, and persona. All sequences are live.
963 live contacts E1 / E2 / E3
4 LinkedIn accounts connected. Each persona covers a specific vertical. Invitation note + post-connect DM + follow-up relance. Pacing: 16 invites/day/account.
Chris — Insurance / FinServ
Jamie — Healthcare
Rodney — pending connection
Sending domains warmed prior to campaign launch. Deliverability monitored throughout to maintain inbox placement for cold outreach at scale.
Warmup complete
Campaign activity, replies, and outcomes synced back to HubSpot for account-level tracking and handoff to the sales team when a contact responds.
Connected
LinkedIn invitation trigger fires automatically when a contact enters the sequence. Post-connect DM fires on connection acceptance — decoupled from email so connected contacts receive a message immediately.
Invitations live Post-connect DM in progress
Outreach across 4 LinkedIn accounts and multiple email senders requires strict rules to avoid over-touching the same company, burning deliverability, or sending conflicting messages.
No more than 5 contacts from the same company touched per week across all channels. For large accounts with 10+ enrolled contacts, the sequence is staggered by tier and buying role priority.
Contacts at the same company receive email from the same sending domain throughout the campaign — avoiding the appearance of a coordinated blast from different senders.
Each contact is assigned to exactly one LinkedIn persona — Insurance/FinServ → Chris, Healthcare → Jamie. The two accounts never contact the same company simultaneously. Assignments are fixed for the campaign lifetime.
16 invitations/day/account maximum. Batch 1 (145 already-connected + pending invites) goes first. Then Tier 1 (763 contacts by score). NorthStar (293) follows after Batch 1 is processed.
183 contacts assigned to Rodney remain assigned — not redistributed. Excluded from all current sends, held until a LinkedIn account is connected. Email fallback considered separately.
LinkedIn DMs trigger on connection acceptance — not tied to email sequence completion. Unblocks ~14 already-connected contacts and ~130 pending invites who were waiting for a message.
Every contact has a fully personalised, AI-written outreach sequence. No shared templates — each sequence is built from the contact's role, their company's confirmed technology landscape, and the most relevant proof point for their vertical.
Built from: the contact's buying role (economic buyer / technical evaluator / champion), the account's confirmed or suspected CCM vendor, the strongest proof point for their vertical, and 2–3 personalisation hooks tied to their title and live signals.
E1 — Problem-first. Pain angle for their specific role, no ask.
E2 — Reframe. A different angle on the same problem, one proof point.
E3 — Proof + soft CTA. Named case study, asset, low-friction ask.
1,492 / 1,493 written
Invitation — ≤250 chars. Relevance-led, zero pitch.
Post-connect — ≤500 chars. Light problem question, signed by persona.
Follow-up — ≤500 chars. One concrete proof point, soft CTA.
In generation — 1,256 contacts