How a person moves through

loading…

1 · Found

you

Every journey starts with a campaign: a name, a goal in plain words, and a target count (Campaigns → new campaign). Creating one starts a running mission — the scout keeps itself in flight from there, picking a strategy (notable people, or people at notable companies) and returning real candidates ranked by how well they fit the goal. Campaigns run in parallel, take turns on the scout, and each only sees the people it found. Duplicates are skipped automatically.

you: Write the goal when you create the campaign. The scout runs itself after that.

2 · Researched

engine

Each new person gets background research from public knowledge: who they are, what they run, likely pain points (labeled by confidence), and conversation hooks. An ICP fit score (0–100) says how well they match your goal. The engine never scrapes private pages or invents facts — unknown stays unknown.

3 · Drafted — with receipts

engine

The drafter writes a first-touch message that must cite at least two research facts. If it can't — or won't — the draft is bounced automatically and rewritten. No generic messages survive.

4 · De-slopped

engine

Every draft is scored by a detector trained on how AI text actually reads: em-dashes, cliché closers, buzzwords, over-even rhythm. Any single tell sends it back for automatic rewrite (max twice). What survives reads like a person typing.

5 · Your call

engine + you

The draft lands in your approvals queue with its evidence attached: research facts cited, voice score, ICP fit, and the full message with research phrases highlighted.

you: Approve, edit inline, or reject — right on the campaign page, no hopping. Pick any earlier version (v1, v2…), or Redraft — which re-runs research from scratch, not just the writing. Add steering notes anytime: they steer every future draft, newest notes hardest.

6 · Sent

engine

An approved message is delivered through the platform itself (LinkedIn or Instagram) at human pace. The moment a send is confirmed, the person moves to Contacted in your funnel.

7 · Followed up (politely)

you

If days pass with no reply, the engine drafts one follow-up at a time — same voice rules, same approval queue, referencing your last message without guilt-tripping. It stops the instant they reply, or after your max.

you: Set follow-up timing in Settings (or turn it off).

8 · They replied

engine + you

Replies are read back in automatically. The contact moves to Engaged instantly. If their message confirms or disproves a researched pain point, the hypothesis auto-checks (or crosses out) — with an audit trail you can review. Mentions of news, funding, or launches trigger auto-research with citations so your reply can reference reality. A draft reply waits in your approvals, weighed by your steering notes.

you: Take the conversation — or just approve the suggested reply. This part is yours.

9 · Closing

you

When talk turns to terms, mark them Negotiating so the funnel reflects reality.

you: Run the call, send the proposal, close.

10 · Won (or Lost, or Nurture)

you

Every ending is recorded: Won, Lost, or Nurture (not now, keep warm). Your funnel shows real conversion at every step — what got here, what didn't.

you: One tap on their card.

always true
  • Nothing sends without your approval — one message at a time.
  • Blocked people can never be messaged, period.
  • Research comes from public knowledge, honestly labeled; unknowns are never invented.
  • Follow-ups stop the moment someone replies.
  • Every action the system takes is logged and visible in Activity.
  • Your steering notes and hypothesis outcomes persist per person — the system learns your read on each lead.

stage markers render live from the funnel — if the pipeline changes, this page follows. copy verified against the running system 2026-08-15. nothing publishes anywhere without your tap.