Real estate
Friction kills lead capture. So after Fotocasa Pro Valencia, I'm daring you to break our real estate AI
Published by
SME Analytica
Date
June 17, 2026
Yesterday I spent the day at Fotocasa Pro Academy in Valencia, the tenth edition. I was in the audience, not on stage.
One line from the program stuck to me all day: "la fricción mata la captación." Friction kills lead capture. It was the actual title of José Luis Echeverría's talk on leadgen, and the same idea kept surfacing in the others. Lledó Museros on using AI to save time without losing control. Ismael Kardoudi asking why agents aren't capturing more. Ángeles Nieto on the quiet crisis of the modern agent. Borja Nicolau on sounding human in a digital world.
The whole program was soaked in AI. ChatGPT Images 2.0 in the Innovation Corner, AI for legal documents, AI as your personal assistant, AI to write content that sells. Everyone in the room wanted the same thing from it: their time back.
But most of the tools on the slides still carried the old tax. Steps. Integrations. Dashboards. Setup. To save an agent ten minutes, you first ask them for an afternoon of configuration and a login they'll forget by Friday. That's the friction the talks kept naming, and it's the reason most of these tools never make it past the pilot.
I left with one conviction. The way to sell this isn't to explain it. It's to let people break it.
So here it is. Try to break it.
Update, 19 August 2026: the public demo line is live again after a number rotation. The challenge stands. Open the demo and try to break it.
No pitch, no deck. Just the agent itself, on a link and a QR code. Scan it and you're in a WhatsApp chat with our real estate agent. No signup. No install. No "message me first and I'll set you up."
Talk to it like a buyer, the most difficult one you can imagine. Ask about the flat. Push back on the price. Change your mind halfway through. Ask for a viewing on Sunday at 11pm. Try to trip it up.
The lead you already lost this week
Here's the part every agent at that event already knows in their gut. You're showing a flat on a Tuesday afternoon. Meanwhile a buyer taps Contact on another of your listings, and the portal sends its pre-filled line: "Hi, I'm interested in this flat and would like to arrange a viewing." Another enquiry arrives as a missed-call notice with nothing but a number. Both sit in the inbox until you're back at your desk, and by then the buyer has written to other listings and booked a viewing with whoever answered first. The flat was fine. The price was fine. The agency was just busy with another buyer.
Reply timing is one part of the workflow, not a conversion guarantee. An agency still needs correct listing context, qualification rules, a valid next step, and a person ready to take over.
What the agent actually does
Open it and you can test the three buyer-conversation languages currently documented: Spanish, English and Arabic. Ask about the sample listing, change your requirements, and see when the demo answers or prepares a human handoff.
Conversa can ask about budget, financing, timing and area before it prepares a viewing step. Calendar progression depends on the configured availability and rules; the agency keeps judgment and confirmation where required.
Why I want agents to try to break it
I mean the challenge literally. Break it, screenshot it, and lunch is on me. I don't think it's flawless. But the fastest way to trust a system is to attack it yourself and watch it hold. A demo I control proves nothing. A demo the skeptic controls proves everything. That's the whole reason it's a link in your hands, not a slide in mine.
The friction budget nobody talks about
Most real estate AI dies in the pilot, and not because the model is bad. It dies because adoption has a friction budget, and these tools spend the whole thing before they send a single reply. Every setup step, every integration, every dashboard an agent has to learn gets paid up front, against a benefit that shows up later. Agents are busy and rational. They don't pay.
The workable version meets the buyer on WhatsApp while keeping the agency's review path visible. It still needs scoped setup: listing context, qualification questions, calendar rules, and a named handoff owner.
What this is
The agent behind that link is Conversa, our WhatsApp AI for real estate teams. It keeps an existing portal, site or email enquiry connected to the property context, buyer-need questions, and the next valid human or viewing step. The current offer is designed for agencies already handling inbound buyer demand; it does not claim response-time, conversion, customer, or market outcomes. See the current scope, publication gates, and limitations.
If you run an agency and you've ever watched a lead go cold while you were out on a viewing, open the demo and try to break it yourself. Play the buyer. See if you catch it out.
And I'm genuinely asking: how long does a portal enquiry usually wait in your inbox while you're out showing a flat?
If you would rather read before you poke at it: what a WhatsApp AI agent does for a Spanish agency, and what to check before buying one, and where it should stop and hand the conversation to a person.
Common questions
How fast should a real estate agency reply to a new lead?
Set a response standard your team can observe, then measure qualified conversations and valid next steps. This page does not claim a specific reply time causes a booking.
Can AI handle real estate leads on WhatsApp?
Conversa can answer initial questions from approved listing context, qualify buyer needs, and prepare a valid viewing step or human handoff. Calendar progression depends on configured rules.
What languages does it handle?
Spanish, English and Arabic, in whichever language the buyer writes. That matters in markets like Valencia and the Costa, where a large share of buyers are international.
Does it replace the agent?
No. It handles the first response, the qualifying questions and the scheduling, so the agent spends their time on viewings and closings.
Share this article
Have a workflow like this one?
We map it with your team, connect it to the tools you already use, and keep it running in production.
Talk to us