AI Won’t Replace the Agent Who Owns the Relationship

By Drew Wilke, Head of Enterprise Sales, Luxury Presence

I sat on a panel last month at the Thrive Broker Summit called The Growth Stack: The Habits, Systems, and Strategies Driving Brokerage Success, moderated by Kevin Van Eck, alongside James Sanders, CEO and broker-owner of Estate Properties International, and James Dwiggins, co-founder of NextHome.

James Dwiggins made a point early on that stuck with me. "Agents don't leave over splits," he said. "They leave because somebody forgot about them."

That's true of clients too, and it's the real risk AI actually poses to this business, distinct from the one most agents are bracing for.

AI threatens transactional value

Ask most agents what scares them about AI and you'll hear a version of the same answer. AI can pull comps, draft a listing description, answer a question about school districts, and generate a CMA in seconds.

If your value to a client is mostly the information you hand them, that's a real threat. AI already delivers information faster and cheaper than any agent can.

But information was never the whole job. The agent who wins is the one a client calls first when something in their life changes: a job offer in another city, a divorce, a growing family, or an inheritance.

AI can answer a question, but it can't sit across the table during the biggest financial decision of someone's life and know what they actually need to hear.

Dwiggins put it plainly: "AI is not going to replace the agent unless the agent loses the relationship with the client." That's the real test.

Agents who built their business on being the only source of information are exposed. Agents who built their business on being the person the client trusts are not.

The 8-to-13-year test

Real estate isn't a subscription business. Most clients only transact every 8 to 13 years, which is a long time to stay relevant to someone who isn't shopping for a home right now. Most agents lose the relationship in the years of silence that follow.

Dwiggins puts it bluntly: "if agents nurture that relationship and really dig deep on it," he said, "there is no AI that's going to replace them."

This is where the brokerage side of the equation comes in. If staying close to clients between transactions is what keeps an agent safe from AI, brokerages need to coach and measure for exactly that, not just for how many agents are on the roster.

Sanders put that into practice. He'd rather run 500 agents who are actually working their sphere than the 700 his brokerage carried under its old national franchise, most of whom weren't.

Over the past 18 months, his team coached out about 150 agents, the ones going quiet on their own clients between deals, and rebuilt around the ones who stay in touch. He calls it coaching out. I call it protecting the brand.

AI can scale the relationship

This is where the AI-versus-agent framing breaks down. The best use of AI in this business is helping agents hold more relationships, better, at the same time.

A great agent can only carry so many people in their head. AI can flag when a client hasn't been touched in six months, remember an anniversary, or draft the first pass of a personal note so the agent isn't starting from a blank page.

Sanders described what that actually looks like day to day. "We know their children's name, their spouse's name," he said.

Used well, AI multiplies the exact behavior that keeps an agent irreplaceable, which is the idea behind our AI CRM, built to give any agent the same growth engine a top producer already has.

That changes what brokerages should be training agents to do. Most of the AI conversation right now is about tools: which CRM, which AI assistant, which transaction platform to adopt.

Fewer brokerages are training agents on the thing that protects them long term: how to build and work a sphere.

Teach the tools without teaching the relationship, and you're preparing agents to compete on the one thing AI already does better.

The search behind the search

There's a second shift happening alongside all this, and it's the one I spend most of my time on. Consumers are increasingly researching real estate through AI before they ever talk to an agent or visit a brokerage site.

Roughly two-thirds of consumers already use AI search somewhere in their home buying journey, and that number is climbing.

For brokerages, that raises the cost of blending in. A brand without a clear point of view gets flattened into a generic result next to every other brokerage in the market.

Your brand used to be a marketing asset. Now it's also a data asset.

What AI models learn about your brokerage, and what they surface when someone asks about agents in your market, is shaped by the content and reputation you've built online. That's already happening.

What brokerages should do next

Brokerages that want to keep growing should do two things at once: coach agents to work their sphere as hard as they work new leads, and build a brand sharp enough to get chosen when a consumer's first stop is an AI search instead of a search bar.

Dwiggins put the fix simply. "Make sure your agents are following up with their clients, building those relationships, nurturing those relationships," he said. "We've been doing this for 40 or 50 years."

Track how often your top agents touch their sphere between transactions. That number will tell you which of them are built to last through the next one.