AI-Powered Lead Qualification for a Real Estate Agency

Real estate agencies live and die by speed-to-lead. A prospect who fills out a contact form on a property listing and doesn’t hear back within an hour is already talking to a competitor. Most agencies know this — and most still rely on a salesperson checking their inbox twice a day.

This case documents how a 12-person real estate agency in Colombia automated their entire lead intake process using no-code tools, with zero code written.

The problem

The agency was receiving 80-120 leads per week across three channels: their website, WhatsApp, and a property portal. Each lead required a manual step: someone had to read the inquiry, assess the budget and intent, assign it to the right agent, and send an initial response.

Average response time: 28 hours. Conversion from lead to first call: under 4%.

The solution

They built a qualification workflow in n8n that runs the moment a new lead arrives:

Step 1 — Unified intake: A webhook captures leads from all three sources and normalizes them into a standard JSON object (name, contact, property interest, budget range, message).

Step 2 — AI scoring: The message and stated budget are sent to GPT-4 with a prompt that scores the lead on three dimensions — intent (browsing vs. ready to buy), budget match (vs. current inventory), and urgency signals in the message. Output is a score (1-10) and a one-sentence summary.

Step 3 — Routing: Leads scoring 7+ are immediately assigned to a senior agent and trigger a WhatsApp message via Twilio: “Hi [name], I just saw your inquiry about [property]. I’m [agent], your dedicated advisor. When’s a good time to talk today?” Leads scoring below 7 go into a nurture sequence.

Step 4 — CRM update: HubSpot is updated automatically with the score, summary, and assigned agent. No manual data entry.

Results

After 60 days:

• Average response time dropped from 28 hours to 4 minutes
• Lead-to-call conversion increased from 4% to 18%
• Senior agents focused exclusively on high-intent leads — their time per deal dropped by 35%
• Setup cost: approximately 40 hours of n8n configuration + $120/month in API and tool costs

What made it work

The key insight wasn’t the AI scoring — it was the WhatsApp response speed. Buyers in competitive markets respond to immediacy. The AI scoring helped route correctly, but the conversion jump came from the sub-5-minute first contact.

The team used n8n self-hosted on a $25/month VPS. The GPT-4 API calls cost under $40/month at their lead volume. Total infrastructure: under $70/month for a system that effectively added a full-time intake coordinator.

Replication notes

This pattern works for any service business with inbound leads and varying intent signals. The core components — unified webhook intake, AI scoring, conditional routing, immediate first contact — transfer directly to legal services, financial advisory, home services, and B2B sales.

The n8n workflow template for this setup is available in the InsiderAITrends templates section.