Summary: Zillow is leaning hard on artificial intelligence to pull itself through a weak housing market. The company is betting on AI across product display, search, and internal productivity. That bet could protect Zillow’s place in the market — or it could hand the front door of home search to Big Tech chatbots. This post parses the risks, the tech, the ethics, and the choices Zillow faces, and it asks the question that matters: how does a platform keep control of the transaction when the interface itself may disappear?
Interrupt — Engage: Zillow has gone wild for AI. That’s the headline. Now ask yourself: will those AI tools lock users in, or will they hand the keys to OpenAI and Google? Will Zillow convert “searches on chatbots instead of using Zillow directly” into Zillow traffic — or watch them walk away?
Market context: a weaker housing market, a battered valuation
Housing activity is soft — 4.1 million existing-home sales last year versus a normal range of 5.5–6 million. Zillow’s revenue picture is mixed: the company reports some growth, but market sentiment has cut the stock to roughly a quarter of its 2021 peak. That gap between operational performance and investor confidence is where strategy has to be sharp. Zillow can talk growth, or it can show defensible moves that preserve the business it already has.
AI as “an ingredient rather than a threat”
Jeremy Wacksman calls AI “an ingredient rather than a threat.” Repeat that: an ingredient rather than a threat. Treating AI this way is the right mental model. It frames the technology as something you mix into products, not a competitor that replaces you overnight. But a model is not a moat. If users start their home searches through large language model chatbots and never click through to Zillow, that ingredient gets consumed on someone else’s platform.
So the question becomes practical: how do you keep that ingredient in your kitchen, not theirs? How do you make Zillow the best place to finish a transaction after an AI sets the table?
Display tech: SkyTour, Gaussian Splatting, and Virtual Staging
Zillow is investing heavily in presenting listings better. SkyTour uses Gaussian Splatting to convert drone footage into high-detail 3D renderings. Virtual Staging fills empty rooms with furniture that never existed in the original photos. These are useful tools: they help people imagine living in a space and can speed decision-making.
But they raise trust issues. When photos are augmented or fabricated, viewers must know what they’re seeing. Wacksman says virtually staged images should be watermarked and disclosed. That’s the right baseline. The tougher task is policing sloppy use and maintaining credibility across millions of listings. Can Zillow ensure the licensed professionals follow the rules? Can it audit at scale? Those operational questions are as important as the rendering tech itself.
Search risk: chatbots as the new front door
Here’s the core vulnerability: listings are public. If a chatbot can pull the same listings, plus a friendly summary, scheduling, and mortgage preapproval, why would a buyer open Zillow? Mirror that phrase: “why would a buyer open Zillow?” That’s the question Zillow needs to answer every day.
There are three paths Zillow can take: 1) become the best content source so chatbots must link to it, 2) embed Zillow services into the chatbot ecosystem, or 3) compete on experience so deep that users prefer Zillow for transactions. Each path has trade-offs. Which one will Zillow commit to?
OpenAI and Google: platform partners and potential rivals
Zillow moved fast to integrate with OpenAI. In tests, chatbots didn’t reliably hand off to Zillow — links bounced among Zillow and rivals. OpenAI’s posture is “we’re learning,” which is code for uncertainty. Google is more worrying: it’s testing native listing displays in search that include staging-like presentations and agent booking functions. If Google folds that into its AI, it can present a home and a booking flow without Zillow ever seeing the customer.
No: Zillow can’t assume brand loyalty will hold when the user interface changes. Wacksman believes the “deep, integrated experience” of Zillow will matter when buyers and sellers enter the transaction. That’s plausible. But platform power is real: if Google or an LLM front end becomes the habitual starting point for home searches, Zillow faces a tough, fast-moving reordering of who owns the buyer intent.
Productivity gains and workforce decisions
Inside Zillow, AI is already productive. Engineers produce more code, support tasks are automated, design cycles shortened. The company kept headcount relatively flat while trimming people described as “not meeting a performance bar.” That phrasing is deliberate. It signals a performance culture, not blanket cuts.
For leaders, this is a double bind: push productivity to improve margins, but don’t hollow out the human relationships that give Zillow its local credibility with agents and buyers. How does Zillow keep the best of both — faster product cycles and human trust — without undermining its network of partners?
Ethics and trust: virtual staging and disclosure
Virtual staging helps sales. But it also invites deception if disclosure is weak. Zillow’s honest move is to watermark and label AI-generated content clearly. That reduces friction and preserves trust. Yet the social pressure to make listings “pop” will push some actors to loosen rules. Zillow must set standards and enforce them, or its valuation will become a casualty of eroded consumer confidence.
Ask yourself: what would you trust more — a cleanly labeled staged photo or one with no notice at all? That answer tells you why disclosure isn’t a compliance checkbox; it’s a product feature.
Strategic choices Zillow should be making
Zillow has options. My shortlist — practical, measurable, and urgent:
- Own the query. Build APIs and partnerships so LLMs have to call Zillow for authoritative listings and enriched data, not scrape competitors. Offer simple activation flows so chatbots hand users to Zillow reliably.
- Make the transaction native. Integrate preapproval, agent messaging, booking, and signing into one cohesive flow. If an AI sets the expectation, Zillow must be the place to finish the job.
- Lock in trust signals. Standardize and enforce disclosure for virtual staging and synthetic media. Public trust is a competitive asset; guard it.
- Monetize provenance, not just eyeballs. Charge for verified listings, enhanced tours, and concierge services that LLMs can’t easily replicate without Zillow’s licensed network.
- Measure attribution differently. Track downstream conversions from LLM-originated queries and price that value into partnerships with platform providers.
Which of these should Zillow commit to first? Which will most defensibly keep users inside Zillow’s transaction funnel? Those are the tactical questions for the next 12 months.
What agents and consumers should watch for
Agents: watch where buyer intent shows up. If buyers arrive through a chatbot, you may need to adapt your lead flow and conversation starters. Consumers: demand clear labeling for anything that’s AI-generated. You have the right to know whether furniture in a photo actually exists.
I’ll mirror that idea: ask for clear labeling. Ask for clear labeling. That simple move protects buyers and keeps market incentives honest.
A pragmatic assessment
Zillow has two assets: a household brand that shaped consumer expectations about home value and a massive data and engineering footprint. It also has a real structural vulnerability: public listings and the rise of AI front ends. Wacksman’s stance is steady and plausible, but steady is not the same as secure.
So here’s a blunt point: No, Zillow can’t sit back and hope the world starts using its app by habit alone. It must be proactive about integrations, standards, and product flows that make Zillow the logical place to finish a home transaction — not just the place users visit for a Zestimate.
Questions to provoke constructive action
If you were on Zillow’s executive team today, what single partnership or product change would you push first? If you’re a real estate agent, what would compel you to prefer Zillow leads over those routed by a chatbot? If you’re a buyer, what disclosure or feature convinces you that Zillow’s virtual staging is honest and useful?
Asking those questions out loud will produce better answers than hoping the platforms pick the “right” outcome for you. How do you answer them?
#Zillow #RealEstateAI #PropTech #AIethics #HomeSearch #SkyTour #VirtualStaging
Featured Image courtesy of Unsplash and Jakub Żerdzicki (hGMONQ4T7Co)
