Qualifying 300 monthly leads without hiring a third salesperson
A regional real estate agency was drowning in inbound enquiries from property listings. We built an AI agent that qualifies leads, answers property questions, and books viewings — 24/7.
The challenge
A regional agency listing 60–80 properties at any given time was receiving 300+ enquiries per month across email, web forms, and their property portal. Two agents were spending half their working week answering repetitive questions — price, availability, nearby schools, transport links — before ever getting to a qualified viewing conversation. Response times were slow, and hot leads were going cold.
What we built
We built a property assistant chatbot embedded on their listings website, backed by a live property database in PostgreSQL that syncs nightly with their internal system. Claude handles the conversation — answering questions about specific properties, comparing options, and qualifying buyer intent through natural dialogue. When a lead is ready to book a viewing, the assistant hands off directly to Calendly for the relevant agent's calendar. All conversation data and lead scores feed back into their CRM via n8n.
The outcome
The AI agent now handles the full first-contact flow for over 300 leads per month. 41% of conversations end in a booked viewing — a conversion rate the agency's best human agents would be proud of. The two agents now spend their time almost entirely on viewings, negotiations, and closing — the work that actually generates commission.
The before state
An agent would arrive at work to 20–30 unread enquiries, most asking the same five questions. Responding took up most of the morning. Afternoon enquiries often didn't get a reply until the next day. Weekends were worse — portal leads from Saturday and Sunday sat unanswered until Monday.
What we built
The assistant has access to live property data — price, bedrooms, EPC rating, tenure, vendor situation — and can answer detailed questions about any listing. It uses Claude's reasoning to handle nuanced comparisons and vague queries like 'something with a garden under 350k near good schools'. The qualification flow is conversational, not a form — buyers don't realise they're being qualified. When intent is confirmed, handoff to Calendly is seamless.
How it runs today
The agency's CRM now shows every lead's AI conversation summary alongside their contact details. Agents walk into viewings already knowing the buyer's situation, timeline, and constraints — because the AI extracted all of that in the chat. The owner is considering expanding the system to cover their lettings side as well.