How AI Decides Which Real Estate Agent to Recommend (And How to Make Sure It's You)

Here's something to think about: your next client might never Google you.

Instead, they'll open ChatGPT, Claude, or Perplexity and type something like, "Who's a good industrial broker in Los Angeles?" A few seconds later, they've got three or four names. If yours isn't on that list, you didn't lose the deal. You just never knew the opportunity existed.

AI is the new referral source a lot of brokers are missing. It isn't based on who knows you or who you've had lunch with. It's based on what the internet says about you.

In LA, the stakes are even higher. It's one of the most competitive brokerage markets in the country, and it isn't really one market. It's dozens of them: the Westside, Downtown, the San Fernando Valley, the South Bay, the San Gabriel Valley, Mid-Wilshire, the Gateway Cities. Clients don't search for "a broker in LA." They search for someone who knows their part of town. AI is built to answer exactly that kind of specific question.

Here's the catch. Nobody outside these AI companies knows the exact formulas, and they change all the time. But the basic mechanics are well understood. So let me walk you through what actually happens when someone asks an AI for a broker, what it's looking for, and what you can start doing about it today.

So how does AI actually find you?

AI pulls from two places, and for "find me a broker" questions, the second one matters most.

The first is what it learned in training. That's basically a giant snapshot of the internet taken months or even years ago. If you were well documented back then, the AI might already know who you are. If you weren't, it doesn't.

The second is a live web search it runs when someone asks the question. These tools rely on that more and more to answer local, current questions like ours. That's good news, because it means what you do online today can show up tomorrow.

It searches like a really fast researcher

Here's roughly what happens when your prospect asks for an industrial broker in the South Bay:

  • The AI breaks the question into a handful of searches, like "industrial broker South Bay Los Angeles," "top industrial brokers Torrance," and "Carson warehouse leasing broker."

  • It reads the top results: brokerage bios, LinkedIn profiles, LoopNet and CoStar listings, news stories, award lists.

  • It looks for the names and facts that keep coming up.

  • It writes a quick summary and recommends a few people.

If you're not on the pages it reads, you can't be in the answer. You can be the best broker in your market, and it won't matter if the AI can't find you.

The six things AI is looking for

The way I think about it, AI is lazy in a good way. It rewards the agents who make its job easy. Be specific, be factual, be consistent, and have other people backing you up.

Let's go through each one.

1. Do you match what they asked?

The AI is matching your pages against the question. Where? What kind of property? What service: leasing, sales, investment, tenant rep, landlord rep?

If your page says "industrial leasing broker serving the South Bay and Gateway Cities," that's a match. If it says "full-service real estate professional," it honestly matches nothing. I see this all the time, and it's one of the easiest things to fix.

What to do:

  • Put your niche in your headline, everywhere. "Retail Investment Sales Broker, Westside Los Angeles" beats "Commercial Real Estate Advisor" every time.

  • Spell out the cities, submarkets, and counties you actually work, such as Santa Monica, Culver City, Burbank, or Downtown. Not just "Southern California," and not just pins on a map.

  • Talk the way your clients talk: "warehouse space," "office lease," "1031 exchange," "NNN investment."

2. Can it pull out the facts?

AI is great at reading clear sentences. It's terrible at reading your logo, your fancy graphics, or the brochure PDF you're proud of.

It wants the basics in plain text: your name, title, firm, markets, specialties, years in the business, deal volume, and a few notable transactions.

What To Do:

  • Write your bio like a fact sheet that happens to be in sentences. For example: "Jane Smith is a Senior Vice President at XYZ Commercial in Los Angeles. She has represented tenants in more than 200 office leases across Century City, Downtown, and Santa Monica since 2011."

  • Move the important information out of images and PDFs and onto the web page itself.

  • List your notable deals with the property type, size, and city.

3. Does your story check out everywhere?

This one's big. When your website, LinkedIn, brokerage bio, LoopNet or CoStar profile, and news mentions all say the same thing, the AI gets confident that you're real and the facts are right. When they don't line up, that confidence drops.

It's just like a client checking you out. If your LinkedIn says one firm and your website says another, they wonder what's going on. So does the AI.

What to do:

  • Pull up every profile you have and make your title, firm, markets, and specialties match exactly.

  • Clean up old brokerage bios and any profiles from firms you've left.

  • Pick one version of your name and stick with it. "Joe," "Joseph," and "J." can look like three different people.

4. Are other people vouching for you?

You already know this from the referral world. Somebody else saying you're great beats you saying you're great every single time. AI works the same way. It looks for reviews, press, deal announcements, awards, association listings, and podcast or article mentions.

What to do:

  • After every closing, ask the client for a Google review or a LinkedIn recommendation. Make it a habit.

  • Send your closed deals to the Los Angeles Business Journal, Bisnow Los Angeles, and The Real Deal. They're often looking for content, though larger deals tend to get more traction.

  • Make sure you're actually listed in the directories of the associations you pay dues to: AIR CRE, CCIM, SIOR, NAIOP SoCal, ICSC.

  • Say yes to podcasts, panels, and reporters looking for a market quote.

5. Do you know your stuff?

This is where you give the AI a reason to point to you as the expert: content that answers real questions in your niche, like market reports, deal breakdowns, videos, and FAQs.

It doesn't have to be fancy. It just has to be useful.

What to do:

  • Put out a short quarterly market update for your submarket and property type, like Westside office, South Bay industrial, or Valley retail.

  • Take the questions clients ask you every week and turn them into an FAQ page.

  • Add transcripts or written summaries to your videos. The AI can't watch you, but it can read what you said.

6. Are you still active?

Recent listings, recent deals, and recent posts all tell the AI you're out there doing business. A bio that hasn't been touched since 2019 looks stale, to the AI and to clients.

What to do:

  • Refresh your bio and deal list at least once a quarter.

  • Post about every new listing and every closing.

  • Find a rhythm you can keep. One post a month beats a burst of activity followed by silence.

Here's what AI isn't doing (and why that's great news)

The AI isn't checking your production numbers against some database. It isn't calling your past clients. It isn't vetting your reputation the way a real referral would. It only knows what's written about you in public.

In a market as crowded as Los Angeles, that's the whole opportunity.

The agent with the clearest, most consistent, best-supported online presence often gets the recommendation over a bigger producer who barely exists online.

If you're newer to this business, that should get you fired up. It levels the playing field. And if you've been in the game 20 years and built everything on relationships, take this as a friendly heads-up. The next generation of clients is asking a machine, and right now that machine might never have heard of you.

Your 30-day game plan

The good news is you can knock most of this out in a month, a few hours a week. Here's how I'd break it up.

Week 1: Find out where you stand

  •  Ask ChatGPT, Claude, Gemini, and Perplexity, "Who are the top [property type] brokers in [your LA submarket]?" See who comes up, and pay attention to why.

  •  Google yourself. Write down every profile and mention on the first two pages.

Week 2: Get your facts straight

  • Rewrite your bio in plain sentences with your niche, markets, years in the business, volume, and a few notable deals.

  • Make your title, firm, markets, and specialties match on your website, LinkedIn, brokerage bio, and LoopNet or CoStar.

  •  Fix or delete the old profiles you forgot about.

Week 3: Get other people talking

  • Ask five recent clients for a review or recommendation.

  • Send one recent closing to the Los Angeles Business Journal, Bisnow Los Angeles, or The Real Deal.

  • Check that your association directory listings are up to date.

Week 4: Show what you know

  •  Publish one market update or deal breakdown in your niche.

  •  Build an FAQ page from the questions you get asked most.

  •  Put a recurring quarterly reminder on your calendar to refresh your bio, deals, and listings.

At the end of the day, AI recommends the brokers it can find, understand, and trust. None of that takes a big marketing budget. It just takes being clear about what you do, consistent about who you are, and visible in the places the AI is looking.


So here's my challenge. Sometime this week, open one of these AI tools and ask who the top broker is in your niche and your corner of LA. If it's not you, that's okay. Now you know exactly what to work on.

If this was helpful, come hang out with me on YouTube, Instagram, and X at @JoeKillinger, and check out more at joekillinger.co. I'm always sharing what's working in commercial real estate.