Zillow dropped a bombshell last week. The company is hosting an AI Summit for Investors. And here's what makes this different from every other corporate press release you've scrolled past today:
Software stocks just lost $200 billion in market cap. Investors are running for the exits on anything with "AI" in the name. Yet Zillow is doubling down.
That's either incredibly brave or incredibly stupid. Let's figure out which.
Why This Matters Now
On February 23, 2026, Zillow Group announced it will host an investor event focused on AI innovation in real estate. The company trades under two tickers: Z (Class A) and ZG (Class C). Same business, different voting rights.
Prediction markets give ZG a 65% bullish probability after the summit. That's not a slam dunk. But it's also not the kind of number you see for a company about to flame out.
The Real Story Behind the Headlines
Here's what the press release won't tell you.
Zillow has 230 million monthly users. That's not a vanity metric. That's raw material. Every search, every saved home, every "zestimate" check feeds their machine learning models.
Think about it this way: Your competitor can copy your features in six months. But can they copy 15 years of transaction data? Can they replicate the behavior patterns of nearly a quarter-billion users?
That's the moat. And it's deeper than most investors realize.
The Numbers That Actually Matter
| Metric | Value | What It Means |
|---|---|---|
| Monthly Active Users | 230M+ | Massive AI training data |
| Zestimate Accuracy | 2.4% median error | Real ML advantage |
| AI Summit Date | TBD (announced Feb 23) | Catalyst timing |
| Software Sector Hit | -$200B selloff | Sector headwind |
| Bullish Probability | 65% | Market positioning |
The 2.4% median error on Zestimates isn't perfect. But it's better than what Redfin or Realtor.com can offer. That accuracy gap compounds over millions of home valuations.
- 230M+ monthly users = unmatched AI training data
- 2.4% Zestimate accuracy creates competitive moat
- Pivot from iBuying to AI-as-a-service reduces risk
- Low expectations from sector selloff = upside surprise potential
- $200B software sector selloff creates headwinds
- Zillow Offers failure still fresh in investor memory
- Redfin, CoStar Group competition intensifying
- AI monetization timeline remains unclear
The Elephant in the Room
Let's address the obvious question. Zillow tried AI-powered home buying before. Remember Zillow Offers? The company lost hundreds of millions when their pricing algorithms couldn't handle market volatility.
They shut it down in 2021. The stock cratered. Critics declared the company dead.
But here's what those critics missed: Zillow didn't abandon AI. They pivoted. Instead of buying homes directly, they're now selling AI tools to the people who do: agents, lenders, and institutional investors.
Same technology. Different business model. Lower risk.
What to Watch at the Summit
If you're thinking about ZG stock, three things matter:
Revenue attachment. Can Zillow show how AI features translate to paid subscriptions or transaction fees? Generic AI hype will get punished. Specific monetization will get rewarded.
Competitive moats. Redfin, Realtor.com, and CoStar Group are all investing in machine learning. Zillow needs to explain why their data advantage is permanent, not temporary.
Profitability timeline. Zillow has always prioritized growth over margins. If they can point to a clear path from AI investment to bottom-line results, the stock moves higher.
Prediction
Direction: Bullish | Probability: 65% | Horizon: 30 days
Zillow's AI Summit is a catalyst event with real upside. The company's data moat is genuine. The pivot from iBuying to AI-as-a-service reduces risk. And the recent sector selloff means expectations are low enough that a solid presentation could surprise to the upside.
The 65% probability reflects legitimate sector headwinds. But it also reflects something more important: Zillow isn't just another AI pretender. They've been doing this for years.
The market will figure that out.
Technical Analysis
365 trading days of data for ZG (2024-09-06 to 2026-02-20)