NeuSo funktioniert es
Alle Neuigkeiten
Verbesserung2 Min. Lesezeit

Your first email, now picked from several drafts

For AI Autopilot first emails, the AI now writes several versions and automatically sends the strongest one — scored by your Self Learning AI and checked for quality before it ever leaves the outbox.

Your first impression in a cold inbox is the whole game. So we taught your AI Autopilot to stop settling for its first attempt.

What changed

Until now, the AI wrote one opening email per prospect and sent it. Good — but a single draft is a single guess. With this update, the AI writes several distinct versions of each first email in one pass: different openers, different structure, different calls to action. Then it picks the strongest one to send. You get the benefit of a writers' room without the wait.

How it picks the winner

The choice isn't random, and it isn't "longest" or "fanciest." Each candidate is scored by your Self Learning AI — the same brain that has been learning what your audience actually replies to. The version that best matches the voice, length, and angle your prospects reward is the one that ships.

Every candidate also has to clear a quality gate before it can win:

  • No slop — generic filler and obvious AI tells are screened out.
  • Reads clean — length and reading level stay tight for a cold inbox.
  • Stays safe — spam-trigger phrasing and broken merge fields are blocked.

If a draft fails any of these checks it can't be selected, so the email that goes out is both the most promising and the cleanest.

Why it matters for your outreach

First emails carry disproportionate weight. A weak opener means no reply, no follow-up sequence gets read, and no meeting gets booked — regardless of how good the rest of your campaign is. More evaluated shots on goal means a better first email on average: sharper openers, tighter copy, and a close that fits your market.

Because the scoring is tied to real outcomes, the bar keeps rising as your campaign sends. Early drafts feed the model, the model picks better drafts, better drafts earn more replies — the quality loop tightens on its own without you touching a setting.

Safe by design

This runs quietly underneath your campaign. If the selection step can't complete for any reason — no clear winner, missing signal, any hiccup — the AI falls straight back to its proven single-draft path. There's no new setting to manage and nothing to break in a live send.

Where it's live

Best-of-several drafting is on for AI Autopilot first emails, where the AI is already drafting and sending for you. Manual and template campaigns are unchanged. It pairs naturally with Self Learning AI: one learns your taste, the other puts that taste to work on every new conversation.

This is rolling out in Beta to everyone now. We're watching closely and reading your replies — tell us what you see.