When someone replies to your outreach, PlacementFlow's AI reads the message and decides what to do next — and the moment any reply arrives, follow-ups to that person stop automatically. This guide explains how classification works, what happens automatically, and what always requires your attention.
PlacementFlow monitors your connected mailbox for incoming replies:
When a reply arrives, the system matches it to the original outreach email using the conversation thread, records it, and pauses the follow-up sequence so nothing else sends while the reply is being handled.
The AI reads the reply text and determines the sender's intent from phrasing, tone, and context. Classification happens automatically within seconds — you don't need to do anything. If the AI's output looks untrustworthy (or the reply is in a language it can't confidently read), the reply is escalated to a human instead of being guessed at.
| Intent | Example | What It Means |
|---|---|---|
| Interested | "Sounds great, let's talk" | Positive response, wants to proceed |
| Not Interested | "No thanks, not for me" | Clear rejection |
| Request for Information | "Can you tell me more about the role?" | Asking a question |
| Out of Office | "I'm away until Monday" | Temporary unavailability |
| Timing | "Not right now, check back next quarter" | Open to future contact |
| Call Request | "Give me a call to discuss" | Wants a phone conversation |
| Counter Offer | "I'd need at least $150K to consider" | Negotiating terms — always human |
| Unsubscribe Request | "Please remove me from your list" | Opt-out request |
| Do Not Contact | "Stop emailing me" | Stronger opt-out |
| Spam Threat | "I'm reporting this as spam" | Complaint threat — escalated and suppressed |
There's one more state the AI never picks itself: Human Required. It's assigned when classification fails or is uncertain — those replies go straight to a human instead of triggering any automation.
| Intent | Automatic Action |
|---|---|
| Interested | The next step is prepared — a screening link if the role has an active questionnaire, or a CV request — and you're notified to take over |
| Not Interested | Candidate excluded from this search, sequence ends |
| Request for Information | The AI drafts an answer and notifies you to review and reply |
| Out of Office | Return date parsed, sequence pauses and resumes when they're back |
| Timing | Candidate parked as dormant and automatically resurfaced later |
| Call Request | You're notified immediately with the context — a human always takes the call |
| Unsubscribe Request | Added to the suppression list, all campaigns to them stop |
| Do Not Contact | Hard block applied, candidate excluded |
| Spam Threat | Address suppressed immediately, urgent alert to you |
Counter-offers always go to a human — the AI never negotiates money. This is a hard rule, regardless of your autonomy level. The classifier is deliberately trigger-happy here: any mention of salary expectations, compensation conditions, or a competing offer is treated as a counter-offer, not as "Interested."
When someone counters, you'll receive an urgent notification and the reply lands in your human-review queue. The AI won't send anything — you decide how to respond.
When the AI can't confidently classify a reply (mixed signals, ambiguous language, an unfamiliar language), it's marked Human Required and routed to you. The sequence stays paused until you've read it and decided what to do. You'll never see the AI silently guess.
If a candidate has opted out of automated decision-making (GDPR Article 22), the AI skips classification entirely for their replies — a human handles them directly.
When someone replies, follow-ups to them stop automatically. This prevents sending "just following up" right after they said yes, emailing them on vacation, or continuing to pitch after a decline. On multi-recipient submissions, only that recipient's chase stops — other recipients continue on schedule.