The buyers and sellers in your market aren’t searching the way they used to. They’re asking. ChatGPT, Perplexity, Gemini, Google’s AI Overview — they’re all answering the same question your next client is typing right now: “Who’s the best agent in [my city]?”
This playbook gives you the four-part system for becoming the agent AI recommends. No ad spend. No new tools. Just the work that gets your name into the citations AI pulls from when it answers — so when somebody asks, you’re the one it talks up to prospective clients.
Real estate has always been a know-you, like-you, trust-you business. The new question is this: does AI know you, like you, trust you? Have you given it everything it needs to talk you up to prospective clients?
This is a Goldilocks moment. Most agents haven’t even heard of this work yet. Get it done now, and you build a recommendation engine that pays you back for years. Wait, and somebody else in your market becomes the answer.
Before the four steps, one thing has to click: AI doesn’t answer from memory. When someone asks ChatGPT or Perplexity or Google AI Overview who the best agent is (or related prompts), the AI runs a live web search, pulls answers from the websites it trusts, and synthesizes a recommendation from those sources — the citations.
Search didn’t go away. It got outsourced. The AI does the searching for the customer, then hands back a clean answer with a few names attached. Your job is to be one of the names.
AI checks two layers when it answers a question:
- Search — it scans search results to see titles, descriptions, and snippets
- Fetch — for the sources that look most useful, it actually visits and reads the page
You want to win at both layers. That’s why steps 2 and 3 are different moves: step 2 makes you look right in search results, step 3 makes you look right when AI reads your page.
One concept ties all four steps together. AI can’t directly read most reviews — they’re locked behind login walls and platform restrictions. Even when AI can see them, it’s lazy. It’s not going to read 800 Zillow reviews to figure out who’s best. Instead, AI grabs high-authority content where somebody else has already done the synthesis: a Reddit thread that names you, a Facebook group comment that recommends you, a “best agents” article that lists you, a profile that signals you’re the obvious choice.
Each step in this playbook is a different way of putting that signal in front of AI. Sites it can crawl. Profiles it can scan. Pages it can cite. Comments it can quote. The work is making sure AI has plenty of high-authority signal pointing back to you.
That happens through four moves. Do them in order.
AI doesn’t invent agents out of thin air. It pulls from a small, predictable set of websites. If your name and brand aren’t on those sites, you’re invisible — no matter how good you are at your job.
Group the sites into four buckets. Lead with the ones you control, then work outward.
Owned — what you control
Listicles — the “best agents” directories
Social — where conversations happen
Local — your community footprint
Do this now
Showing up on the right sites is the table stakes. The next move is making your profile on each of those sites the one AI picks over your competitors. This is profile engineering — writing every bio, headline, and “about” section so AI reads you as the obvious answer.
AI is doing pattern recognition. It scans hundreds of agent profiles and surfaces the ones with the clearest, most specific signals. Vague profiles lose. Specific profiles win.
What to engineer into every profile
Want the full bio rewrite system? AiM has a dedicated tool for this. Optimize Your Bio for AI Search Engines walks you through the exact prompt and process for engineering each profile.
Do this now
Profiles get you on the radar. Proof pages put you in the answer. A proof page is a piece of content on your own website designed to get cited when AI runs a “best agent” search.
AI loves listicles, comparisons, and answer-style content. It rewards pages that say BEST, TOP, and VS. Most agents don’t publish anything close to this — which means the lift to dominate isn’t huge.
What proof pages look like
Three sample headlines, all built for AI citation:
Notice the pattern: city name, clear answer-shape, specific year or number where it makes sense. AI grabs these because they map cleanly to the questions buyers and sellers are typing.
How to write them so AI cites them
Proof pages are even more powerful when they target a specific niche — luxury, relocation, military, expireds, move-ups, FTHBs, and more. AiM has a dedicated guide for this. Specialization Pages for Answer Engines shows you exactly which niche pages to build and how to structure them.
Do this now
You don’t have to be the top producer in your market to be the agent AI recommends. The underdog angles — Reddit, Facebook groups, YouTube, Instagram, LinkedIn — are where smaller agents get cited alongside the big names. Sometimes ahead of them.
Each one works a little differently and rewards a different kind of presence. Pick the ones that match how you already show up — then show up better than the agents who ignore these channels entirely.
Reddit (and Facebook groups)
Lead with Reddit. It’s lower friction than Facebook groups — public, no membership barrier, and AI cites it heavily. Facebook groups can work too, but most are private or require approval, which makes the path from “find the thread” to “your client comments in it” a heavier lift.
The threads getting cited are often a year or two old — they don’t need to be fresh. Volume of comments and depth of substance matter more than recency.
The Google hack for finding local threads
Use this Google search to surface every local Reddit and Facebook thread where people are asking for agent recommendations:
The keyword “recommendation” is doing real work here — it’s the exact word AI is also looking for when it scans these platforms. Try variants: “recommend,” “who would you recommend,” “any good realtors.”
How to work the threads
YouTube — and how AI actually reads videos
YouTube is the single most-cited website across all AI searches. But here’s the part most agents miss: AI doesn’t watch your video. It reads the text layers around it — the title, the description, the chapter timestamps, and the transcript.
Listing tour videos rarely get cited because none of those text layers are answering anything. Talking-head explainer videos do get cited — when the title, description, transcript, and chapters all line up around a clear question and a clear answer.
How to make AI cite your videos
Instagram and LinkedIn
Instagram is starting to get cited too — and AI is reading the comments, not just the captions. Encourage real conversation under your posts. Keywords in comments are showing up in citations, which means a comment thread full of clients naming your city and your wins is doing real work.
LinkedIn articles are also breaking through as proof-page substitutes (see step 3). If you don’t have a website, LinkedIn is your fastest path to publishable, citable content under your own name.
Do this now
Tips
From the Session: This playbook was introduced in AI Recommendation Engine.