Search just had its biggest overhaul in a quarter century, and most real estate agents don’t know they’ve already lost ground because of it.
That was the warning from Jeff Lobb, founder and CEO of SparkTank Media and Coach52, during a session at Inman Connect San Diego on Wednesday.
Google’s recent pivot to AI-driven search — folding more query traffic into Gemini’s conversational answers — marks the company’s most significant search change in 25 years, according to Lobb. And it’s rewriting how consumers find real estate agents, lenders, title companies and attorneys before an agent ever knows they were in the running.
“When people ask me where they should spend their time or money, I ask them: Where’s your customer? Are they on social?” Lobb said. “Because right now, they’re shifting to AI significantly, the same way we saw the shift from the Yellow Pages to Google.”
Blue links are out. Direct answers are in
The shift Lobb described isn’t just Google indexing pages differently. It’s a change in how people search at all. Where search used to mean typing a few keywords into a box and clicking through blue links, Lobb said it’s increasingly a conversation. Consumers ask a direct question, get a direct answer and, often, never click through to a website at all.
That’s fueling the rise of what’s being called AEO — Answer Engine Optimization, sometimes AI Engine Optimization — as a companion discipline to traditional SEO.
Lobb cited industry estimates of a 20 percent to 40 percent drop-off in traditional search-and-click behavior as AI answers absorb queries that used to generate a page visit.
Buyers and sellers, he said, are increasingly asking AI platforms directly for home valuations, agent recommendations and referrals to attorneys or lenders. These are the same queries that used to route through Zillow, Redfin or a Google search.
Lobb didn’t cite a specific percentage of consumers doing this, but argued the bigger problem is on the supply side. The vast majority of agents, he said, aren’t optimized to be found in that kind of search at all.
“You can look at that as bad news — more work — or as an opportunity,” Lobb said. “If you get ahead of it now, you won’t be the person 5 years from now saying, ‘I wish I’d started this sooner.’ We missed that boat with Google. This is the new one.”
When AI cites your worst review
Part of what makes AI search different from traditional SEO, according to Lobb, is that large language models don’t just rank pages but read and synthesize language, including the specific wording in online reviews. AI can then surface it as reasoning in an answer.
Lobb described asking an AI platform, on a call with a brokerage, which firms were among the top in Brooklyn. The brokerage he was working with appeared in the results. But the AI, unprompted, offered to walk through the pros and cons of each firm listed.
One “con” it surfaced: a negative characterization of the brokerage’s agents as “creepy” drawn from a four-year-old Yelp review, unrelated to the brokerage’s current standing, that AI had judged credible enough to cite.
“I had nothing to say to that. AI can and does read reviews, and it will use the language from them in its answers,” Lobb said. “Different models pull from different sources.”
Lobb also pointed to industry estimates that put the share of AI-generated, inauthentic-sounding reviews at just under 25 percent. It’s a trend he attributed to businesses trying to keep pace with rising review counts rather than any coordinated effort.
Consumers, he said, generally can’t distinguish AI-written reviews from real ones, but the practice muddies a data source that AI models increasingly treat as an authoritative signal.
Search yourself as a stranger would
Lobb’s core recommendation was less about gaming any particular platform and more about a content strategy built around how consumers actually phrase questions, not how agents assume they do.
He suggested agents write down the specific questions a buyer or seller in their market might ask an AI platform (“Who’s the best luxury real estate agent in Tampa to sell my waterfront condo?”) and build content such as blog posts, video and social posts that directly answer each one.
He recommended two free tools for gauging exposure: Searchably.ai, his own company’s tool, which scores a website’s visibility for a given search term on a 0-to-100 scale, and AnswerThePublic.com, a longer-standing tool that surfaces the actual questions consumers are typing or asking in a given market.
Lobb also urged agents to test their own visibility the way a real consumer would: logged out, or in an incognito browser window, rather than through an account that AI platforms may already associate with the agent’s own content.
“It might make you mad,” he said, “because you might not be the name that comes up.”
The channels AI is actually reading
Once a business identifies its target questions, Lobb’s suggested workflow runs through several channels.
Publish the answer as a blog post on a website capable of being crawled by AI systems, then post a version to a Google Business Profile, which he said carries more weight now that it feeds directly into Gemini’s results.
From there, he recommended repurposing the same answer as graphics-based content for LinkedIn, Instagram and Facebook, plus a short video, ideally under 90 seconds, addressing the question directly on camera. Lobb specifically flagged YouTube Shorts as underused given its ownership by Google and its likely role as a Gemini data source.
He also named two platforms he said are punching above their weight in AI search results right now.
FastExpert was one, which he attributed partly to a lack of competing content in the space. And Reddit, which he said can work in an agent’s favor only with active monitoring, since unmanaged threads can surface negative or inaccurate characterizations without an agent’s knowledge.
Lobb was clear that none of this requires an SEO background — he said he doesn’t have one himself — but does require agents to think and search the way a consumer does rather than the way an industry insider does.
“We’re all using AI daily, and buyers and sellers are, too, but most of our marketing hasn’t caught up,” Lobb said. “If someone is asking AI who to hire right now and you’re not part of that answer, you have no way of knowing you lost that opportunity. That’s the real risk.”
