Lead Qualification & Scheduling Bot

Business Challenge

Solstice Realty's agents were losing leads to slower response times — inbound inquiries from listing sites often waited hours for a reply, by which point many prospects had already contacted a competing agency.

Objectives

• Respond to every new inbound lead within a minute, day or night
• Qualify leads (budget, timeline, financing status) before handing off to an agent
• Book showings directly into agents' calendars without back-and-forth

Approach

As a five-person brokerage, Solstice needed something they could run without a dedicated ops person — so we deliberately kept the system's configuration simple enough for the founder to adjust qualifying questions herself post-launch.

Deliverables

An SMS and web-chat lead qualifier, direct Calendly integration for showing bookings, and a weekly summary email of qualified vs. unqualified leads.

Solution

Planning

We scoped a narrow first version — qualify and book, nothing else — rather than the fuller CRM-replacement Solstice initially asked for, since the immediate pain was response time, not their existing CRM.

Design

Conversation flow mirrors how their best agent naturally qualifies a lead over the phone, based on call recordings the team shared during discovery.

Development

Built on Next.js with the OpenAI API for conversation handling, Twilio for SMS, and a direct Calendly API integration so bookings land on agents' real calendars with no manual step.

Testing

Two weeks of live shadow testing where the bot drafted responses that a team member manually approved before sending, to catch any qualifying-question issues before going fully autonomous.

Deployment

Launched fully autonomous after the shadow period showed consistent, accurate qualification with no approval overrides needed for five consecutive days.

Optimization

We tuned the qualifying questions once after launch, dropping a financing question that leads found intrusive early in the conversation and moving it later in the flow — booking completion rate improved as a result.