Skip to content
‹ Back to Agent Factory
LEAD_CAPTURE

Booking & follow-up

Qualified leads get to a calendar, get reminded before the call, get re-engaged when they ghost, and get nudged back to life when they stall. The pipeline stops leaking between "interested" and "booked."

Overview

The gap between "qualified lead" and "meeting on the calendar" is where most pipelines quietly lose more deals than the discovery call ever does. The lead asks for time. The rep replies. The lead doesn't reply back. A week passes. The rep moves on. The deal that never was gets logged as "no response" and the team blames the lead instead of the follow-up.

This agent owns the entire stretch from "I'd like to talk" to "you showed up to the call." It proposes times, books the meeting, sends the reminders, follows up when prospects go quiet, re-engages no-shows without the awkward chase, and nudges stalled leads back into motion. Reps stop chasing replies. The pipeline stops leaking. The meetings that should happen actually happen.

Capabilities

  • Proposes meeting times tied to the assigned rep's real calendar availability, with the prospect's timezone respected from the first message
  • Sends reminder sequences before every booked call, 24 hours, 1 hour, 15 minutes, tuned for the meeting type and the prospect's behavior
  • Follows up automatically when a prospect goes silent, with a multi-touch cadence that backs off appropriately and doesn't read like a script
  • Re-engages no-shows the same day with a tone that recovers the relationship instead of burning it
  • Nudges stalled leads, qualified, interested, but inactive for 5+ days, with context-aware outreach that references the original conversation
  • Tracks every meeting through its full status lifecycle (proposed, booked, confirmed, attended, rescheduled, no-show, recovered) and reports the funnel back to RevOps

Example Output

Trigger

Qualified lead Maya Chen replied "yes, I'd like to set up a call" to AE Jordan Rivera at 4:32pm Wednesday. Jordan is on a customer call until 5pm.

Agent action, sent at 4:33pm, from Jordan Rivera's address:

Hi Maya,

Glad we're getting this on the calendar. I've got a few times that should work for both of us, accounting for the fact that you're on Pacific time:

  • Thursday 11:00am PT
  • Thursday 2:30pm PT
  • Friday 9:30am PT

Pick whichever works and I'll send the calendar invite. If none of those land, here's my full availability: [link]

Looking forward to it,

Jordan

Then, downstream: Maya picks Thursday 11am PT. Calendar invite goes out within seconds. 24 hours before the call, a reminder lands in Maya's inbox with the meeting agenda Jordan typically uses for first-call discovery. One hour before the call, a Slack DM pings Jordan with a one-paragraph brief on Maya and Trellis Health. 15 minutes before the call, Maya gets a final reminder with the Zoom link.

No-show case (alternative path): Maya doesn't show up. At 11:18am PT, the agent sends a recovery message:

Hi Maya, looks like something came up. No worries, it happens. Want to grab a different time? Here's my calendar: [link]. Or if the timing isn't right anymore, just let me know and I'll close the loop.

Stall case (alternative path): Maya replied initially but never picked a time. Six days pass. The agent sends a context-aware nudge:

Hi Maya, circling back on getting time on the calendar. Last week you mentioned you were evaluating against [stated alternative] and trying to make a call before end of quarter. Happy to keep this on track if it's still a priority. If the timing's shifted, just let me know.

Logged to CRM: Meeting status, reminder cadence, no-show recovery attempts, stall-nudge sequence, all written to the lead record with full timestamp history.

Agent Workflow

Step 01

Detect the booking signal

The agent triggers on positive booking intent: a reply containing meeting language, a calendar link click, a "find time" form submission, or an explicit handoff from another agent. The signal type determines the response register.

Step 02

Propose times in the prospect's timezone

The agent reads the assigned rep's real calendar availability, applies meeting-length rules and buffer preferences, and proposes 2-3 options in the prospect's local timezone. No "let me know what works", the prospect picks from a curated list, with a fallback link if none fit.

Step 03

Book and confirm

Once the prospect picks a time, the agent books the meeting on the rep's calendar, sends the calendar invite with the right video link, and writes the meeting record to the CRM. Confirmation goes to the prospect within seconds, not minutes.

Step 04

Run the reminder cadence

The agent sends a tuned reminder sequence: 24 hours, 1 hour, 15 minutes. Each reminder carries the right context, the meeting agenda, the video link, the dial-in if needed, and the cadence shifts based on meeting type and the prospect's prior reliability.

Step 05

Brief the rep before the call

One hour before the meeting, the assigned rep gets a Slack DM with a short brief: who the prospect is, what company, why they reached out, what's been said so far, what's worth opening the call with. The rep walks in already oriented.

Step 06

Handle no-shows the same day

If the prospect doesn't attend, the agent sends a recovery message within 20 minutes, short, gracious, with a fresh booking link. No guilt, no friction, no awkward chase from the rep the next morning. Most no-shows reschedule on the recovery message.

Step 07

Nudge stalled leads

For qualified leads who replied but never booked, the agent runs a multi-touch nudge cadence over the following 7-14 days. Each message references the original conversation context, what the prospect said, what they were evaluating, what timeline they mentioned, so the outreach reads as continuation, not pursuit.

Step 08

Track status and report the funnel

Every meeting moves through its full status lifecycle, every status change writes to the CRM, and the agent surfaces the funnel back to RevOps: proposal-to-booking rate, booking-to-attended rate, no-show recovery rate, stall-revival rate. The leak points become visible.

Go live in days

AI for work, that just works.