SalesMind AI treats every message as part of a relationship, not an isolated blast: connection requests start fresh, follow-up touchpoints build on what was already said, and a reply stops the automated sequence so a human conversation can take over.
What does each stage know about the past?
| Stage | What the AI uses |
|---|---|
| Connection request | The prospect's profile and activity — no prior thread exists yet |
| Follow-up touchpoints | Earlier messages in the sequence — later touchpoints can reference the angle already used, instead of repeating it |
| After a reply | The sequence stops for that lead; the conversation moves to your inbox (and Autopilot, if enabled) with full context preserved |
💡 This is why follow-ups do not repeat your first message: each touchpoint is written with awareness of what the prospect already received, so the sequence reads as one coherent conversation rather than four disconnected pitches.
What happens when a prospect replies?
The automated sequence for that lead ends. Replies belong to a different job — answering the person — which is handled in your inbox with the conversation history intact. Interest, objections and questions deserve a response to what the person actually said, not the next timer-driven message. See Start Using the New Autopilot & CRM Features.
What about negative signals?
Learning respects negative outcomes: conversations where a prospect opted out or responded negatively are excluded from the examples the AI learns from. Your thumbs up / thumbs down feedback sharpens this further — see Train the AI with Thumbs Up / Thumbs Down Feedback.