Cold email fails when it reads cold. Lists get scraped, first lines get templated, and follow-ups stop after one try because a human got busy.
This is where per-agent inboxes matter. Every Squad teammate has its own real email address with its own sending history. One teammate runs a careful sequence from one address. Add teammates and each runs its own lane from its own inbox, so volume grows without one address carrying everything.
Before each send, the teammate researches the prospect with its search tools and its own browser: their site, their recent posts, what they ship. The first line is about them because the teammate actually looked. Replies come back to the same inbox, and the interested ones become flagged tickets in your queue.
Who you sell to and why they care. The teammate builds the list with its research tools and enriches every row before anything sends.
First touch, two follow-ups, a close. Edit the drafts in the ticket; the voice you leave is the voice every send keeps. The squad's Voice file holds the rules.
Before each email the teammate reads the prospect's site in its own browser and checks recent activity with AI Search. No research, no send.
A real inbox with a real thread history and a daily cap you set. Bounces and unsubscribes are honored instantly and logged. More volume = more teammates, each with its own inbox and cap.
The sequence runs on the schedule, per prospect. A reply stops the sequence for that thread the moment it arrives.
A positive reply becomes a ticket with the thread, the research, and a suggested next step. The lead lands in HubSpot enriched. You take the call; the teammate books it.
From: Miles <miles@yourcompany.squad.so> To: priit@ferrostack.io Subject: your Kubernetes cost post Hi Priit, Your post on cutting Ferrostack's cluster bill by 40% made the rounds here. The part about idle staging nodes is exactly the problem we solve for teams your size — automatically, not with a spreadsheet review every quarter. Worth 15 minutes next week? If not, the post was still great. Miles
Your teammate connects through the integrations you approve, works on its own computer with its own browser, and asks you before anything irreversible.
You describe the target and approve the sequence. The teammate builds the first list overnight.
Ten researched drafts wait as tickets. You read the first lines: they cite real posts and real pages. You approve eight.
First replies thread back to the teammate's inbox. Two are interested; both reach you as flagged tickets with the research attached.
You add a second teammate for a second segment — its own inbox, its own cap, its own lane. Volume doubles; quality does not drop.
An AI teammate enriches every new signup the moment it lands: company, role, size, fit, and files it where your team actually looks..
An AI teammate sorts the overnight inbox each morning, answers the routine messages, and flags the few that need you.
An AI teammate spots accounts going quiet each week and drafts the check-in before they cancel.