The problem: rankings held, leads didn’t
Tom has been licensed for nine years. He ranks on page one for his core keywords. He spends $1,500 a month between Zillow Premier Agent and Google Ads. By every traditional metric, his digital marketing is working.
But his leads are down 25% year-over-year. The Zillow leads he does get are lower quality — more tire-kickers, more renters, more people who already have an agent. His Google Ads cost-per-lead keeps climbing even though his click-through rate hasn’t changed much.
Tom’s SEO dashboard looks fine. His rankings didn’t drop. His buyers moved.
This isn’t hypothetical. It’s a pattern showing up across real estate markets nationwide. Agents with stable or improving Google rankings are reporting fewer inbound leads, lower lead quality from portals, and rising acquisition costs. The instinct is to spend more on what used to work — more Zillow budget, more Google Ads spend, more SEO content. But the problem isn’t that those channels broke. The problem is that buyers found new ones.
Three places your leads are leaking
There isn’t one cause. There are three, and they compound:
- Google’s AI Overviews are absorbing clicks from the organic results you rank in. Your listing is still there — but a growing share of searchers never scroll past the AI-generated answer at the top.
- Zillow, Redfin, and Realtor.com launched ChatGPT apps in late 2025 and early 2026. Buyers who start in ChatGPT now search listings without ever reaching Google — or your website.
- Zillow launched AI Mode on March 25, 2026 — a conversational search experience that keeps buyers on Zillow longer, reducing the chance they leave to find a local agent independently.
Each of these shifts individually would dent your lead flow. Together, they explain the gap between “my rankings are fine” and “my phone isn’t ringing.”
AI Overviews: your #1 ranking now gets 58% fewer clicks
Google’s AI Overviews — the AI-generated summaries that appear above organic results — now trigger on 48% of all tracked search queries, a 58% increase year-over-year (BrightEdge, February 2026). For real estate informational queries like “best neighborhoods in Austin for families” or “how much house can I afford,” AI Overviews are especially prevalent.
The impact on clicks is severe:
| Metric | Without AI Overview | With AI Overview | Change |
|---|---|---|---|
| Organic CTR (all positions) | 1.76% | 0.61% | −61% |
| Position #1 CTR | 28% | ~12% | −58% |
| Zero-click rate | ~60% | 80–83% | +33–38% |
| Paid CTR | 19.7% | 6.34% | −68% |
Sources: Seer Interactive (September 2025); BrightEdge (February 2026). Mobile zero-click rate: 77%.
Read that again: 4 out of 5 users who see an AI Overview never click any result. If you rank #1 for “real estate agent [your city]” and Google shows an AI Overview for that query, your effective visibility just dropped by more than half — even though your ranking didn’t move.
This is the disconnect Tom is feeling. His SEO tool says he’s #1. Google Search Console says his impressions are stable or rising. But his clicks and leads are declining because the impressions are increasingly happening in a format where nobody clicks.
Zillow AI Mode and the portal lock-in problem
On March 25, 2026, Zillow launched Zillow AI Mode — a conversational search experience where buyers can chat naturally about their housing needs. A buyer can ask “Find similar homes within my budget that are closer to light rail” and get personalized results without ever leaving Zillow.
This matters for agents because it deepens the portal lock-in problem:
- Before AI Mode: A buyer browses Zillow listings, maybe clicks out to an agent’s website, maybe contacts the listing agent directly. There were exit ramps.
- After AI Mode: The buyer stays in a conversation with Zillow’s AI. It remembers preferences across sessions. It guides them from browsing to scheduling tours to connecting with a Zillow-selected agent. The exit ramps are gone.
Zillow AI Mode is designed to keep buyers inside Zillow’s ecosystem from first search to agent connection. That agent connection flows through Zillow’s Premier Agent program — which means the buyer never discovers agents independently.
Meanwhile, the portals are also expanding into ChatGPT itself:
| Portal | ChatGPT App Launch | What It Does |
|---|---|---|
| Zillow | October 2025 | Listing search inside ChatGPT, routes buyers back to Zillow |
| Redfin | February 2026 | Property search and neighborhood exploration via ChatGPT |
| Realtor.com | March 2026 | Pre-search affordability and neighborhood Q&A, routes to agents on Realtor.com |
| Local agents | — | No presence in ChatGPT |
When a buyer asks ChatGPT “Show me 3-bedroom homes under $400k near good schools,” ChatGPT can now pull live listings from Zillow, Redfin, or Realtor.com. The buyer gets results, clicks through to the portal, and connects with a portal-selected agent. Your website — even if it ranks #1 on Google for that exact query — was never in the picture.
Buyers are asking ChatGPT instead of Google
The shift to AI isn’t theoretical. The numbers are already large:
- 48% of prospective buyers plan to use AI in their home search in 2026 (NerdWallet Home Buyer Report, 2026).
- 39% of prospective buyers are already using AI tools, up 5 percentage points from the previous quarter (Veterans United, Q2 2025).
- 67% of homebuyers using AI rely on ChatGPT as their primary tool for housing data (Florida Realtors / survey data, 2025).
- 54% use Google Gemini for real estate insights (same survey).
- 82% of Americans use AI tools like ChatGPT and Gemini for housing-related information (survey data, 2025).
What are they using AI for? Estimating monthly payments (41%), comparing lender reviews (32%), checking property values (35%), virtual home tours (36%), and learning about housing trends (34%). These are all activities that used to start with a Google search — and lead to clicks on agent websites.
The critical question for agents: when a buyer in your market asks ChatGPT “Who is the best real estate agent in [your city]?” — do you appear in the answer?
For the vast majority of agents, the answer is no. AI chatbots recommend the same 4–5 national portals in over 90% of responses. Local agents appear in fewer than 1% of AI recommendations unless they have exceptionally strong third-party citation density and web presence.
The Zillow spend trap
Tom spends roughly $1,000/month on Zillow Premier Agent. He’s not alone — Zillow’s Premier Agent program generated approximately $1.95 billion in revenue in 2023 (Zillow public filings). That’s billions of dollars agents pay to appear on someone else’s platform.
The complaints are well-documented and getting worse:
- Non-exclusive leads. Unless a buyer proactively opts in, leads are shared with multiple agents. One agent reported spending over $36,000 only to discover this.
- Low conversion rates. Most high-end agents report conversion rates under 2%. Leads often include renters, people who already have an agent, or buyers looking for the listing agent.
- Rising costs. A single non-converting lead can cost $2,500. Agents in competitive markets burn through $50,000+ annually.
- AI Mode deepens the problem. Now Zillow’s AI is the first touchpoint, and it routes buyers to Premier Agent subscribers — making agents even more dependent on the platform.
Meanwhile, Zillow is also under legal pressure. In 2025, Zillow was hit with eight major lawsuits from competitors, regulators, and consumers, covering antitrust, copyright, and RESPA claims. The platform that agents depend on for leads is simultaneously the platform making those leads more expensive, lower quality, and legally contested.
The trap: Zillow’s AI Mode makes buyers more dependent on Zillow. That makes agents more dependent on Premier Agent. That gives Zillow more pricing power. The cycle reinforces itself — unless agents build independent visibility in the AI channels that are growing fastest.
SEO vs. AEO: why you need both
Traditional SEO optimizes for Google’s search results. It still matters — 88% of buyers still use an agent (NAR 2025), and many still find agents through Google. But SEO alone no longer covers the full discovery surface.
AEO (Answer Engine Optimization) is the practice of optimizing your content to be cited by AI systems — ChatGPT, Perplexity, Google AI Overviews, Gemini, and Claude. The core difference:
| SEO | AEO | |
|---|---|---|
| Optimizes for | Google keyword rankings | AI citations and zero-click influence |
| Success metric | Clicks and rankings | AI mention rate, citation share |
| Content style | Keyword-optimized pages | Data-rich, statistically cited, entity-clear |
| Paid option | Google Ads | None — must be earned |
| Local agent opportunity | Moderate — can rank locally | Low today — but winnable with the right content |
The Princeton/Georgia Tech GEO study found that content with statistical citations and clear factual claims was up to 40% more likely to be cited by generative AI systems. For agents, this means your typical “About Me” page and listing descriptions are invisible to AI. But a local market report with specific data points — median prices, days on market, school ratings, neighborhood demographics — is exactly the type of content AI extracts and cites.
Agents who build AEO-ready content now are positioning themselves for the channel that’s growing fastest. Those who don’t are watching their discovery surface shrink as AI absorbs more of the buyer journey.
What to do about it
The leads didn’t disappear. They moved. Fixing the problem means showing up where buyers are actually looking. Here’s what the data says works:
1. Audit your AI visibility
Before you change anything, you need to know what AI currently says about you. When a buyer asks ChatGPT, Perplexity, or Gemini about real estate agents in your market, what happens? Are you mentioned? Are the facts right? A Metricus AI visibility report answers this with data across every major AI platform, not guesswork.
2. Publish data-rich local content
AI systems cite content with statistics, specific data points, and factual claims. Monthly market reports for your ZIP codes — with median prices, inventory levels, days on market, price-per-square-foot trends — give AI something to extract and attribute to you. This is the content gap between agents who appear in AI responses and agents who don’t.
3. Build your third-party citation density
AI learns from the entire web, not just your website. Reviews on Google Business Profile, a complete Zillow agent profile, mentions in local news coverage, contributions to industry publications, discussion presence on Reddit and real estate forums — these are the signals that push you above the AI recommendation threshold.
4. Add structured data markup
Schema.org markup (LocalBusiness, RealEstateAgent, FAQPage) helps AI systems understand what your business does, where you operate, and what you specialize in. It’s one of the simplest AEO wins and most agent sites don’t have it.
5. Re-evaluate your portal spend
If you’re spending $1,500/month on Zillow and Google Ads and seeing declining ROI, consider redirecting a portion toward building your own AI-visible content ecosystem. A $50 Google Ads lead that you own outperforms a $50 Zillow lead shared with three competitors — and content that earns AI citations keeps working without monthly spend.
The bottom line: You can’t buy your way into a ChatGPT recommendation. You have to earn it. The agents who start building AI visibility now — while their competitors are still only optimizing for Google — will have a compounding advantage as more of the buyer journey moves to AI.
This article gives you the framework. A Metricus report gives you the specific data: what AI says about you across every platform, every factual error with its source, and prioritized actions for your market. One-time purchase from $99. No subscription required.
Sources: NerdWallet 2026 Home Buyer Report; Veterans United AI Homebuying Survey (Q2 2025); NAR 2025 Profile of Home Buyers and Sellers; Seer Interactive AI Overview CTR Study (September 2025); BrightEdge AI Overview Coverage Data (February 2026); Zillow Group public filings (2023); Zillow AI Mode press release (March 25, 2026); Redfin ChatGPT app launch (February 2026); Realtor.com ChatGPT app launch (March 2026); Princeton/Georgia Tech GEO study (Aggarwal et al., 2023). AI mention rates based on Metricus internal testing across the major AI platforms (2026). Learn more about how AI visibility scores work.