Lofty took the main stage at Inman Connect San Diego on July 28 to launch House.ai, a consumer-facing product it calls “financial intelligence for homeownership.” Not a CRM upgrade. Not another listing-description writer. A tool that briefs your buyer on their own buying power before you ever hear their voice. Lofty’s pitch is that it delivers “transaction-ready clients” to agents in its network. Read that phrase twice. The client arrives already educated, and somebody else did the educating.
Two days later, Inman itself announced it is building AI tools with Lofty: Ask Inman, a research assistant trained on Inman’s editorial archive and live market data, plus an expanded AI certification program that Inman leadership described as a driver’s license for agents. Both roll out by January. The direction of the week was unmistakable. The vendors stopped selling software to you and started briefing your client.
The first call changed. Your script did not.
Here is the measured gap. A SparkToro analysis of Similarweb data found 68 percent of US Google searches ended without a click in early 2026. When an AI Overview appears, that jumps to 83 percent. Inside Google’s dedicated AI Mode, it hits 93 percent. Google says AI Mode passed one billion monthly users, with query volume more than doubling quarter over quarter. The research phase of a home purchase now happens inside a chat window, and no click ever reaches your site.
The portals moved to meet buyers there. Redfin shipped its ChatGPT app on February 6. Realtor.com followed on March 30. Zillow turned on its own AI Mode for 5 percent of users and told investors on the Q1 call that it drives “deeper, more substantive conversations than in traditional search.” Zillow’s stated goal is buyers who arrive “more informed” so agents spend less time on education.
Translate that from investor language. The first-call value you built a career on, walking a nervous buyer through budgets, taxes, neighborhoods, and process, has been strip-mined by machines that answer instantly at 11pm. A buyer who calls you in August 2026 already has a budget from an affordability model, comps from AI Mode, and a neighborhood read from ChatGPT. Reciting data they already have is worth nothing. Correcting the data they have is worth a commission. Because a lot of what the AI told them is stale, averaged, or flat wrong about your specific market.
What actually launched at Inman Connect, and what it means for a team of one
The launch list from San Diego matters because of who each product serves. House.ai serves the consumer first: it models buying power and financial readiness, then routes a “transaction-ready” lead to a Lofty-network agent. Ask Inman serves the agent with research on demand. Social Studio generates hyperlocal content from Inman’s archive. Inman also unveiled Inman 2.0, a multi-year plan to become what its CEO called the Bloomberg of residential real estate: less publication, more data platform.
Zillow is running the same play at portal scale. Zillow Pro, its connected operating system for agents, goes to nationwide rollout in the second half of 2026 after a beta with 12,000 agents, with Follow Up Boss as the AI workflow engine underneath. Zillow reports Q2 earnings this week, guiding $750 million to $765 million in revenue, and analysts are openly framing the print as a test of whether its AI bets convert.
For a solo agent or a two-person team, the read is simple. The platforms are packaging intelligence for both sides of the transaction, and you will not out-spend them. You can out-local them. Every one of these systems averages across markets and lags closed data. None of them walked your farm neighborhood last week. That is the only edge left, and it happens to be the one buyers will pay for.
The counter-brief: 90 minutes a week that beats the averages
Stop treating AI-educated buyers as a threat and start auditing what educated them. Run this workflow once a week. It costs your time plus a $20 ChatGPT Plus seat, and nothing else.
Step 1: Run your buyer’s queries before your buyer does (30 minutes). Open ChatGPT and Google AI Mode side by side. Paste this prompt, filled in for your two most active neighborhoods:
“I’m considering buying a home in [neighborhood, city]. Give me current price ranges, property tax rates, insurance costs, HOA trends, and anything a buyer should be cautious about. Then list the questions I should ask a local agent.”
Then run the query that decides whether you exist: “Who are the best real estate agents in [city] for [your niche]?” If your name does not appear, you have a second project. Today’s is the first one.
Step 2: Log what the machines got wrong (20 minutes). You are hunting for stale comps, outdated tax rates, insurance numbers that predate this year’s renewals, and inventory claims that miss the last two weeks. In most markets you will find three to five real errors per neighborhood. Write them down with the correct figure and your source.
Step 3: Build the counter-brief (30 minutes). One page, titled “What the AI won’t tell you about [neighborhood], August 2026.” Five bullets: what the models say, what is actually true, and the number that proves it. Your name, license, and date at the bottom. Save it as a PDF and post it as a page on your site.
Step 4: Wire it into intake (10 minutes). Add one question to your lead follow-up: “What did your research turn up so far?” Most buyers in 2026 will name an AI tool. Do not argue with the machine. Send the counter-brief within the hour. A saved template in Follow Up Boss or Lofty CRM makes this a two-click move. You have now repositioned yourself from information vendor to the person who checks the machine’s work.
Step 5: Feed the machines your version. Publish the counter-brief page monthly and mark it up with FAQPage schema, with your bio page carrying RealEstateAgent and Person schema. The systems answering next month’s buyer questions pull from structured, consistent, recently updated sources. Google’s grounding now connects its models to more than 250 million verified places. Make your data the version they trust.
Watch Wednesday’s Zillow print
The 30-day signal is Zillow’s Q2 earnings call this week. Ignore the revenue number. Listen for two operational details: whether AI Mode expands past the 5 percent test group, and whether Zillow Pro’s national rollout gets a firm date. If AI Mode goes wide this fall, the pre-briefed buyer stops being an early-adopter quirk and becomes your default Tuesday phone call.
The certification wave is worth a calendar note too. When Inman’s AI driver’s license program lands in January, it will become a listing-presentation talking point within a quarter. Being early costs a weekend. Being late costs credibility in front of sellers who have started asking what AI tools their agent runs.
The move this month is not buying anything from the Inman Connect launch list. It is running the counter-brief workflow before your next listing appointment, so that when a seller or buyer quotes a chatbot at your kitchen-table meeting, you can pull out one page that shows exactly where the machine was wrong about their street. The agents who lose to AI will be the ones who tried to compete with its speed. The ones who win will charge for its blind spots.
