How Zillow Turned Google NotebookLM into an AI-Powered Marketing Channel

The Problem Zillow Was Trying to Solve

Zillow has spent 20 years building one of the most comprehensive libraries of home-buying content on the internet. Guides, articles, checklists — the kind of deep, trustworthy information that first-time buyers desperately need.

The challenge? Consumers are no longer starting their research on brand websites. They’re opening ChatGPT, asking Google’s AI Mode, and expecting instant, personalized answers. Zillow needed to follow that behavior shift — not just optimize for it on their own domain.

“We’ve got 20 years of articles about how to buy a home,” said Rico Farmer, senior director of brand marketing at Zillow. “The idea of being able to put that all into NotebookLM to create a custom way for our users to interact with it was something we wanted to experiment with.”

The goal wasn’t just traffic. It was presence in AI environments where buyers were already spending time.

Why NotebookLM? The Strategic Logic

NotebookLM isn’t a household name the way ChatGPT is. But that’s partly what made it interesting to Zillow.

The platform is purpose-built for deep research. Users upload source materials, and NotebookLM generates answers, summaries, and interactive content exclusively from those materials — no web scraping, no hallucinations from random corners of the internet. For a brand that cares about accuracy and trust, that’s a significant advantage.

It also maps perfectly to the first-time homebuyer mindset. These are people who read everything before making a decision. They compare, they question, they want to understand every step. NotebookLM is literally designed for that kind of intensive information consumption.

Add in Zillow’s pre-existing relationship with Google, and the partnership made practical sense too.

Inside the Notebook: What Zillow Actually Built

Zillow’s branded notebook is called “Your Ultimate Guide to Homebuying.” It’s trained on 40 of Zillow’s home-buying guides and built around three core interaction modes.

The Brand Safety Angle Marketers Should Notice

One of the quieter but more important details here is what NotebookLM doesn’t do.

Unlike AI search platforms that pull from the open web, NotebookLM only responds based on the materials it’s been trained on. That means Zillow’s notebook won’t surface a competitor’s listing, a misleading Reddit thread, or an outdated news article. Every answer traces back to content Zillow controls.

For brand marketers who’ve been nervous about AI platforms distorting their message, this is a meaningful distinction. It’s a closed, brand-safe environment — and Zillow can update the source materials at any time by swapping guides in and out.

How Rare Is This? The Competitive Landscape

Zillow is one of only a handful of brands that have built a public-facing NotebookLM experience for consumer engagement. Most of the others are publishers — Yahoo, The Atlantic — organizations that already think in terms of content as product.

Google only features around 30 such activations on its “Featured notebooks” page. That’s a remarkably small number given the platform’s 30 million monthly users.

This is early-mover territory. The brands that figure out NotebookLM now are building institutional knowledge about a channel that most of their competitors haven’t even considered yet.

What This Means for AI Marketing Strategy

Zillow’s experiment isn’t just a cool product story. It’s a signal about where brand marketing is heading.

As consumers increasingly rely on AI tools to start their research journeys, the question for marketers shifts from “How do we rank on Google?” to “Where are our customers gathering information, and are we present there?” The answer is no longer a single platform.

NotebookLM represents a specific type of AI user — someone doing serious, intentional research. That’s a high-intent audience. And Zillow is the only real estate brand currently showing up for them in that environment.

“There are so many [AI] tools, and it’s valuable to us to get in early and build our knowledge about them to understand if [they can] deliver something to our customers in a unique way,” Farmer said.

That mindset — test early, learn fast, build channel knowledge before the crowd arrives — is the real takeaway here.

The Limitations Worth Acknowledging

NotebookLM is not a mass-market channel. Thirty million monthly users sounds significant, but it’s a fraction of the audiences on ChatGPT or Google Search. The platform also skews toward research-oriented, tech-comfortable users — which fits the first-time homebuyer profile reasonably well, but won’t reach every segment of Zillow’s audience.

There’s also the question of discoverability. Users have to find Zillow’s notebook to use it. Unlike SEO or paid media, there’s no established distribution engine pushing people toward branded notebooks. Zillow’s ability to leverage its existing Google relationship helped here, but most brands won’t have that advantage.

And as with any early-stage channel experiment, the long-term ROI is still unproven. Zillow is treating this as a learning investment — which is the right framing — but it’s not yet a scalable acquisition channel.

The Takeaway

Zillow didn’t wait for NotebookLM to become mainstream before showing up. They identified a high-intent audience, matched it to a platform built for deep research, and used two decades of content to create something genuinely useful — not just another branded chatbot.

The broader lesson is straightforward: the brands that win in AI-driven marketing won’t be the ones who optimize hardest for today’s dominant platforms. They’ll be the ones who recognize emerging AI environments early, test with intention, and build the institutional knowledge to move fast when those channels scale.

NotebookLM may or may not become a major marketing channel. But Zillow will know the answer before anyone else does.