Zillow’s AI Bulldozer Is Coming — And It’s Aimed at the Process, Not the Product

I saw this article about Zillow’s CTO, David Beitel, and it immediately struck me as having many of the components I’ve tried to highlight in post after post.

We talk about AI. We debate models. We argue about whether the ROI is “real” yet. But what stuck out to me reading this was simpler:

If Zillow can pull off what they’re planning — and a big part of it simply isn’t possible without the current and future versions of AI — what are the first-order and second-order consequences?

Because the real story here isn’t “Zillow adds a chatbot.” The story is: AI becomes the operating layer across the entire home-buying workflow. And once you change the workflow, you change the industry.

AI as a System, Not a Feature

Beitel’s “north star” for two decades has been to infuse technology into every stage of the home-buying process — search, find an agent, schedule tours, make an offer, and complete the transaction. And what’s different now is that he’s describing AI as something you use upfront, not as an afterthought. That’s a strategic tell. When leaders start talking that way, they’re not thinking “feature.” They’re thinking “system.”

What also jumped out: Zillow is treating AI internally the way winning teams treat performance analytics — measure it, test it, iterate, and don’t get sentimental about the tools.

Their engineering teams have been using AI coding assistants (Cursor and Claude) for over two years, and they’re not waving their hands about “vibes.” They’re tracking productivity and quality, constantly swapping models and dialing in the workflow. That’s how you turn AI from a toy into a compounding advantage.

And they’ve pushed adoption beyond engineering. Every employee has access to ChatGPT. They’re working with Glean to let people explore gen AI in a “guardrailed” environment, with permissioning controls, and employees have already built 4,600+ internal AI agents. That number matters. It means the organization is shifting from “a few AI pilots” to “distributed automation,” where the long tail of small workflow improvements starts to add up to real leverage.

The AI Footprint Consumers Already Touch

Now, layer that onto the external customer experience.

Zillow’s AI footprint isn’t hypothetical. It’s already embedded in the pieces consumers touch: Zestimate (AI/ML-driven home valuation), computer vision for 3D walkthroughs, AI-enabled virtual staging, drone-based SkyTour for exterior and neighborhood context, and conversational search so users can ask questions in natural language instead of contorting themselves into dropdown filters.

They’ve also rolled out “AI Assist” for renters — answering practical questions like whether utilities are included or whether parking exists — and they even shipped a ChatGPT integration in roughly six weeks that lets users ask for listings conversationally and then routes them back to Zillow’s app/site. That’s not just distribution. That’s a signal that the front door to real estate search may increasingly be conversational interfaces — and Zillow intends to be present there.

First-Order Consequences: From Hunting to Guided Execution

So what happens if they keep going “deeper into the transaction,” as Beitel describes?

If you’re a first-time homebuyer, the process is daunting because it’s not one decision — it’s a chain of decisions and handoffs, wrapped in anxiety, time pressure, and paperwork. Beitel even calls out how fraught this purchase is for most people.

AI doesn’t just make search prettier. It can:

Reduce cognitive load (natural language Q&A instead of filters)
Increase clarity (visualization tools, staging, 3D walkthroughs)
Reduce friction (scheduling, follow-up, coordination)
And, if they succeed at the “tail end,” compress the most painful part: documents, approvals, and multi-party orchestration

The consumer impact isn’t merely “faster.” It’s less uncertainty per step. And when uncertainty drops, conversion rises — for Zillow, for agents, for everyone in the chain.

Agents Who Adopt Become Super-Agents

Zillow didn’t just focus on the buyer. They bought Follow Up Boss (CRM for agents) for $400M plus a potential earnout, and the description is telling: AI to summarize calls, propose to-do lists, and draft personalized messages.

That’s basically “agent throughput software.”

In a world where leads are expensive and attention is scarce, the winners won’t be the agents with the biggest hustle. They’ll be the agents with the best system. AI turns responsiveness, personalization, and follow-up into something you can scale without burning out.

But that also means the gap widens. If two agents start with the same book of business and one is running an AI-augmented workflow while the other is working manually, the outcome over 12–24 months won’t be subtle.

Second-Order Consequences: Whoever Owns the Workflow Captures the Margin

Here’s the part most people miss when they talk about “AI in real estate.”

The home-buying industry isn’t one market. It’s a relay race of loosely connected markets — portals, agents, lenders, inspectors, appraisers, title/escrow, closing. Each handoff is a chance for delay, misunderstanding, rework, or failure.

If AI can smooth the “tail-end” transaction stages — the tedious coordination of mortgage documents and the endless back-and-forth between parties — then Zillow isn’t just improving user experience. Zillow is moving from “search + leads” toward transaction orchestration.

And when you orchestrate the workflow, you get to decide:

Where demand gets routed
Who gets surfaced
Which services are bundled
And which steps become commodities

This is the bulldozer part.

Because the competition stops being “another listing site.” The competition becomes: anyone who sits inside the workflow today and charges for friction.

Who Wins, Who Loses?

If Zillow executes, the likely winners are:

Platforms that control proprietary data + user intent + workflow execution
Service providers who integrate into that workflow cleanly
Professionals (agents, lenders, etc.) who become trusted advisors instead of human middleware

The likely losers are:

Point solutions that don’t own distribution or workflow
Businesses whose margins rely on manual coordination and opaque steps
Anyone who thinks “AI” means bolting a chatbot onto last year’s process

And there’s a subtler “loser” category too: companies that keep their old organizational cadence while the tech underneath them compounds monthly. Zillow is explicitly describing a culture of constant model/tool testing, plus company-wide access and guardrails. That’s an adoption curve advantage, not a marketing line.

The Industry-Level Change: Consumer Expectations Reset

This is how industries flip.

Not when the technology exists — but when customer expectations reset around a new default.

Once buyers can ask a system, conversationally, “Show me apartments in the West Village under $5,000, and tell me if parking is available,” and get answers immediately, they stop tolerating the old experience.

Once scheduling and follow-up become automated, consumers stop accepting “I’ll get back to you tomorrow.”

Once the document treadmill starts getting compressed, the expectation becomes: “Why does closing take this long?”

AI doesn’t just accelerate the process. It makes the old process feel ridiculous.

Zoom Out: This Isn’t Just Real Estate

Now zoom out: this isn’t just real estate — it’s what happens when AI hits complex, regulated workflows.

This is the part that should make every healthcare executive, payer, provider, and digital health founder sit up straight.

Healthcare is a thicket of multi-party coordination: referrals, prior auth, scheduling, documentation, coding, billing, care coordination, follow-ups. It’s “tail-end transaction complexity” on steroids.

If Zillow can use AI to reduce friction across a long, document-heavy process with multiple stakeholders — while putting guardrails in place so employees only access permitted data — the pattern is portable.

In healthcare, the stakes are higher and the constraints are tighter. But the opportunity is enormous for the same reason: most of the cost and patient dissatisfaction lives in the workflow.

The Core Strategic Takeaway

And that’s the core strategic takeaway I keep coming back to:

AI’s biggest ROI doesn’t come from making individuals faster. It comes from redesigning the workflow so the work changes shape.

So if you’re in real estate, the question isn’t “Do we have an AI strategy?”

The question is: Are we redesigning the workflow before someone else redesigns it for us?

Because if Zillow (or anyone) pulls this off at scale, and your “competitive advantage” is built on yesterday’s friction…

…you may end up standing in front of the bulldozer.

Source: Fortune interview with Zillow CTO David Beitel (Feb 18, 2026).