A real estate team turned $1,400 of ad spend into a $6M close. The repeatable part isn't the ad.

A recent case study: a brokerage paired a modest lead-gen campaign with AI-powered follow-up. Over three months it sent 726 nurture messages across 96 serious prospects, and one of them turned into a $6 million sale. The ad spend behind it was $1,400.

It's a great number. It's also the wrong thing to copy. The ad didn't close the deal. The follow-up did. So here's how the follow-up actually worked.

How it works

The mechanism is simpler than the result makes it sound. Five moves, running on their own:

1. Catch every inquiry the second it lands. A new lead from any channel triggers a reply right away, day or night. This is the part that matters most: leads contacted within five minutes are roughly 21 times more likely to qualify than leads contacted within thirty, and the first business to reply usually wins. No human team can be first every time, across every hour and every channel. The AI can.

2. Open with something real, not a form. The first message answers the question the buyer actually asked, with real detail about a unit, a price, or a timeline. It reads like a person who was paying attention, because it was.

3. Keep the thread warm, on a cadence. Most leads don't buy on day one. They go quiet. The system follows up over days and weeks, in the buyer's language, each message shaped by what the buyer said last, not a generic blast. That's where the 726 messages came from: steady, relevant nudges nobody had to remember to send.

4. Score and route. As the buyer responds, the system reads how ready they are and hands the warm ones to a human with the full history attached. The rep starts the call already knowing the budget, the preference, and the promise made on Tuesday.

5. Write it all down. Every message, answer, and signal lands in one record, so nothing depends on a person's memory and nothing falls through.

That's it. No magic, no special campaign. Every inquiry got a fast, real answer, and nobody got forgotten.

You can do this for your own workflows too

Follow-up is the obvious one, but it's the same five moves everywhere the work depends on someone remembering. Qualifying a raw inquiry into a brief your rep can use. Answering a buyer at midnight with real detail about a unit. Chasing the payment and the document that slip after the sale. Each is a job that leaks money when it runs on memory, and each one an AI can own end to end.

And it compounds. When the same record sits under all of it, every reply, every close, every lost deal teaches the system what works for your buyers. The team that runs this for six months has something the team starting today does not: a book that gets sharper every week.

The part the winners get right

By 2024, about 72% of agents reported using at least one AI-powered tool. The ones actually closing more deals weren't using it to replace the relationship. They were using it to hold the relationship at a scale a person can't: remembering every client, every preference, every promise, and bringing a human in at the moment it mattered.

That's the difference between a chatbot and a colleague. A chatbot answers and forgets. A colleague remembers. The first loses you the $6M deal three months later when the follow-up goes cold. The second is what turned a $1,400 ad into a close.

Where Sonzai fits

Sonzai is the relational intelligence layer for real estate. It keeps a living record of every buyer, deal, decision, and promise, and puts an AI workforce on top of it that runs those five moves around the clock, on the systems you already use. You don't need to build any of it. You point it at the job that leaks the most, and it runs.

If you want to see what your own pipeline looks like with nothing falling through, book a demo. Bring your messiest process. You'll leave with a clear picture of the outcome, on your numbers.