ChatGPT Sent Me a Client: What Actually Made AI Recommend Me

Table of Contents

1. What Happened: ChatGPT Sent Me a Client
2. The Exact Question: A Person, Not an Agency
3. Why the Bar in Digital Marketing Is So Low
4. Why This Is Bigger Than One Lead
5. How AI Changes the Referral Game
6. AI Referrals Act Like Human Referrals
7. Faceless Agency vs. Identifiable Expert
8. Why I Think I Showed Up in ChatGPT
9. The Content I’ve Been Publishing
10. Consistency Creates Confidence
11. Do YouTube Videos Matter for AI Recommendations?
12. Your Biggest Upgrade
13. How to Get Recommended by ChatGPT
14. How to Test Whether AI Recommends You
15. Ask Every New Lead One Question
16. FAQ

What Happened: ChatGPT Sent Me a Client

A couple of days ago, ChatGPT sent me a client. I’m still pumped about it, and I want to walk you through exactly what happened and what I think actually made the difference, because this is where the lead game is heading for every small business.

Here’s the story. Someone in Los Angeles wasn’t happy with the person running their Google Ads. That person had started cranking out junk AI content for them, what a lot of us in the industry now call AI slop. And here’s the thing about AI slop: humans don’t respond to it. Google doesn’t reward it. And the AI tools themselves don’t recommend businesses that are built on it.

I want to be clear about one thing, because it gets lost in these conversations. AI isn’t the problem. I use AI every day in my business. If you use AI to produce genuinely good, helpful content, it’s incredibly useful. The problem is what you produce with it. Lazy, generic, mass-produced content is the problem, whether a human or a machine wrote it.

So this person was frustrated, and they did what more and more people do now. Instead of Googling, they opened ChatGPT and asked it for help. ChatGPT told them, in effect, “If you’re looking for help with Google Ads, you should reach out to Jeremy Ashburn with PushLeads. He’ll help you.”

I’m not in LA. I don’t have any outreach happening in LA. I do have a couple of clients out there, but none of them referred this person. There was no ad, no cold email, no networking connection. An AI model looked at everything it could find about people who help with Google Ads and decided I was the right answer.

That’s exciting. But the more useful question is why it happened, because if I understand why, I can do it again, and so can you.

Section summary: A prospect 2,000 miles away found me because ChatGPT recommended me by name, with no ads, outreach, or local referral involved.

The Exact Question: A Person, Not an Agency

When I got on the phone with this prospect, I asked what they typed into ChatGPT. Their exact question was:

“Who would be the best person to help me set up a Google Ads campaign?”

Notice the word person. They didn’t ask for the best Google Ads agency. They didn’t ask for the top PPC firms. They asked for a person.

That one word changes what an AI model goes looking for. To answer a question about a person, ChatGPT has to find an actual individual who is clearly and repeatedly connected to that skill. A company logo and a “we” on a services page doesn’t give it much to work with. A real name, a real face, and a body of work tied to that name does.

This is a big deal for small business owners, because a lot of your customers are asking the same kind of question. They want the best plumber, the best roofer, the best accountant, the best therapist. They want a person they can trust. If the internet doesn’t clearly connect your name to your expertise, you’re hard to recommend.

Section summary: The prospect asked for a “person,” which favors a clearly identifiable expert over a faceless company.

Why the Bar in Digital Marketing Is So Low

There’s another reason people are asking for a person instead of an agency, and it’s not pretty. A lot of agencies out there don’t do the work.

I regularly talk with business owners who are paying an agency something like $2,000 a month and being told, “Don’t worry, we’ll handle everything.” Then I log in to audit the account, and no work is being done. No new content. No campaign changes. No real optimization. Just a monthly invoice and a report that looks busy.

That’s been my experience with a big chunk of the agencies I audit, and it’s why the bar in my industry is so low. When so many players are doing the minimum, doing high-quality work stands out fast.

Here’s what I’ve found actually gets rewarded, by clients, by Google, and now by AI tools:

• High-quality work. Content and campaigns that actually help the customer, not filler.
• Transparency. Showing clients exactly what was done, instead of hiding behind vague reports.
• Showing up. Meeting with clients, walking them through the work, and answering their questions.

The industry rewards that. It doesn’t reward lazy. And AI models, which are trying hard to give accurate answers, seem to be learning the same lesson.

If you’ve been burned by an agency before, our guide to what good Google Ads management services actually include will help you spot the difference.

Section summary: Many agencies do little real work, so transparent, high-quality work stands out to clients, search engines, and AI.

Why This Is Bigger Than One Lead

One lead is one lead. What it represents is much bigger.

OpenAI reported that ChatGPT reached 900 million weekly active users in February 2026, more than double the 400 million it reported a year earlier. Independent tracking suggests it’s already pushing toward the next big number: according to Sensor Tower data reported by Reuters, the ChatGPT app crossed 1 billion monthly active users in June 2026.

That’s hundreds of millions of people every week asking an AI for answers, and a growing share of those questions sound exactly like the one my prospect asked. Who should I hire? Who’s the best at this? Who can help me?

For a small business, that’s a new referral channel, and it doesn’t care about your ad budget. It cares about whether you’re the most credible answer it can find.

Section summary: With around 900 million weekly users, ChatGPT has become a major referral source, and it rewards credibility, not ad spend.

How AI Changes the Referral Game

For years, local marketing was about rankings. Get to the top of page one. Get into the map pack. Don’t end up on page two, because nobody goes to page two.

AI changes that picture in three important ways.

1. There’s no page two

A Google search shows ten blue links, a map pack, ads, and a pile of other features. ChatGPT usually names two or three businesses, sometimes just one. If you’re not one of those names, you don’t exist for that conversation. There’s no scrolling down to find you.

2. Geography matters less when the question is about expertise

My prospect didn’t ask for someone local. They asked for the best person. That opened the door for someone in North Carolina to win a client in California. For local services like plumbing or roofing, location still matters a lot. But when the question is about skill, AI will cast a much wider net.

3. It rewards giving, not pitching

The way to show up in AI answers isn’t to shout louder. It’s to publish high-quality educational content that genuinely helps people, without asking for anything in return. That’s the kind of material AI tools can quote, cite, and trust.

If you want to see how this plays out for contractors specifically, read what decides whether ChatGPT recommends your contracting business.

Section summary: AI names very few businesses, reaches beyond local boundaries for expertise questions, and rewards genuinely helpful content.

AI Referrals Act Like Human Referrals

Here’s something I didn’t expect. On the same day I spoke with this prospect, I also got a referral from a real person here in Asheville, a local connection who sent someone my way.

Both of those referrals acted the same. The AI referral behaved exactly like someone a trusted friend had sent to me. They came in warm, open, and already leaning toward working with me.

Of course, they were also curious. The prospect told me, “I want to know exactly what you’re doing, why you’re getting AI referrals.” So I walked them through the things I’ll share with you below. I didn’t ask for money up front. I just showed them the value.

That’s a principle I learned from BNI: givers gain. You give without expecting anything in return, and it comes back to you. It’s how I run my business, and it’s how I think about content too.

Then I made them a simple offer. I told them I could start small, managing their Google Ads for $250 a month, and show them what I can do. Ultimately, to really move the needle, most businesses need to invest somewhere around $1,000 to $2,000 a month across everything, including ad spend. But starting small lets me prove myself with consistent, high-quality work. That’s what people are looking for: consistency and quality. I’ll say it again because it matters so much. Consistency.

Pre-sold leads

When an AI recommends you, the lead arrives pre-sold. They’ve essentially been told, “This is the person you should talk to,” by a source they trust. The first call feels less like a sales pitch and more like a confirmation. That’s the most exciting part of this for me. It’s all about the pre-sold lead.

Section summary: The AI referral behaved just like a human referral, arriving warm and pre-sold, and a small, value-first offer made it easy to say yes.

Faceless Agency vs. Identifiable Expert

Let’s go back to that word “person,” because I think it’s one of the most important lessons in this whole story.

Here’s how a faceless agency and an identifiable expert look to an AI model trying to answer “who’s the best person to help me?”

If you were an AI trying to confidently recommend a person, which column would you pick from?

That doesn’t mean agencies can’t win. PushLeads is an agency. But ChatGPT didn’t recommend “some agency.” It recommended a real person who works at a real agency, because that person had a trail of evidence attached to his name.

If your customers hire you personally, whether you’re a solo practitioner, a founder, or the face of a family business, make sure the internet knows who you are.

Section summary: AI can confidently recommend a named expert with reviews, a story, and content, far more easily than a faceless brand.

Why I Think I Showed Up in ChatGPT

I want to be honest here: nobody outside OpenAI knows exactly how ChatGPT picks its recommendations. This is my best read. But I know what changed in my business over the last few months, and I know what I found when I searched my own footprint the way an AI would. It comes down to four components.

1. Pages that directly answer the buyer’s question

I publish pages that answer the exact questions buyers ask. What’s the best way to set up Google Ads? How do you set up your own Google Ads campaign? What should you look for in an agency? I give all of that information away for free. If someone reads it and still wants help, I’m available.

When ChatGPT searches the web to answer “who should help me with Google Ads,” pages shaped like clear answers are exactly what it pulls. A good example is our breakdown of the Google Ads strategy we use for restoration companies.

2. Third-party reviews that describe my Google Ads work

You need signals from other websites, not just your own. My reviews on third-party sites describe the actual Google Ads work I do: campaign setup, optimization, reporting, and results like more calls. An AI trusts what other people say about you more than what you say about yourself.

This is also why this video is on YouTube and LinkedIn and this article is on my website. Your expertise has to show up in multiple places. Reviews matter so much that we wrote about why businesses with too few reviews can be invisible to ChatGPT.

3. A consistent business with history back to 2012

PushLeads has been around since 2012. My website, LinkedIn, directory listings, and profiles all describe the same business with the same services. That consistency over time is a trust signal. There’s no confusion about who I am or what I do.

4. More than 100 YouTube videos

I have more than 100 videos on my YouTube channel, and the big difference in the last couple of months is that I’ve been producing a lot more content. More on that in the next section.

Section summary: Direct-answer pages, third-party reviews, a long consistent history, and a large video library likely combined to make me a confident answer.

The Content I’ve Been Publishing

My YouTube channel has a little under 700 subscribers right now. That’s not a big number, and I don’t care. I don’t measure success by subscribers or views. I measure it by how many high-quality videos I’ve published that answer real questions.

That’s an important mindset shift. If you judge your videos by view counts, you’ll probably quit. A video with 40 views that answers one exact question can still be the thing an AI reads, trusts, and uses to recommend you.

I’m publishing three types of videos:

• Slideshow videos. I introduce myself, answer a specific question, and walk through a lot of data on screen. These aren’t slop. Each one is built around a real question and a real answer.
• Talking videos with graphics and B-roll. The real version of me, talking directly to the camera, with visuals to support the points.
• Loom-style videos. Quick screen recordings where I answer a question someone asked, like this one.

Every one of them focuses on answering questions my customers actually have. That’s the common thread.

Section summary: I publish three kinds of question-answering videos and judge them by quality and usefulness, not views or subscribers.

Consistency Creates Confidence

This brings me to what I think is the most important idea in this article: consistency generates confidence.

AI models cross-check things. Yes, AI can still hallucinate. But the companies building these tools are working hard to prevent that, and the models are getting better at avoiding it. One of the main ways they avoid mistakes is by checking whether multiple independent sources agree.

So when an AI sees your website with helpful articles and videos, then sees your LinkedIn, then sees your directory listings and reviews, and they all tell the same story, the model starts treating that story as settled fact. It’s a kind of triangulation. Several points of evidence, all pointing to the same answer.

That’s where the confidence comes from. That’s why ChatGPT didn’t hedge with a list of ten random agencies. It said, essentially, “I highly recommend Jeremy with PushLeads.” Not a nameless, faceless agency, but a real person at a real agency, backed up by consistent evidence.

The opposite is also true. If your business name is different on different sites, your services are described inconsistently, or your website says one thing while your reviews say another, AI gets less confident. And a less confident AI recommends someone else. We cover the most common gaps in why your business doesn’t show up on ChatGPT.

Section summary: When your website, profiles, reviews, and content all agree, AI treats your expertise as fact and recommends you with confidence.

Do YouTube Videos Matter for AI Recommendations?

I was curious about this myself. My assumption was that ChatGPT doesn’t really care about videos. And in one sense that’s true: ChatGPT doesn’t watch your video.

But it can read everything around it: the title, the description, and the transcript. And every one of my videos is also published on my website with the title, the full transcript, the chapters, and video schema. So every video is another page tying my name to my expertise.

Think of it as a library. For a person, my YouTube channel is a library they can browse to decide whether I know what I’m talking about. For an AI, it’s the same thing: a huge library of evidence showing that I help people, answer real questions, and don’t ask for anything in return. That adds up to a very favorable picture.

When I started doing this, I thought I’d mainly show up in Google’s AI, meaning Gemini and Google’s AI Overviews, which rely heavily on YouTube. That’s a huge part of showing up in AI. We’ve seen it in our own data too: in our AI citation study across 45 businesses, YouTube was one of the most cited sources in AI answers.

What surprised me was ChatGPT finding me. My best explanation is that it’s reading the transcripts on my site, seeing that I publish a lot of helpful content, and seeing that the content consistently connects me to Google Ads and SEO.

I’ve also broken down the bigger picture in a separate video on why AI keeps citing YouTube.

Section summary: ChatGPT doesn’t watch videos, but it reads titles, descriptions, and on-site transcripts, which turns a video library into a large body of readable evidence.

Your Biggest Upgrade

So what should you actually do with all this? Here’s the single biggest upgrade most small businesses can make:

Make videos that answer your customers’ most common questions, and publish them on your website as text.

A few specifics that matter:

• Go long enough to be useful. My rule of thumb is 10 to 20 minutes. In my experience, the longer, more complete videos are the ones AI tools pay attention to and cite. Short clips are great for social media, but the in-depth answer is what earns trust.
• Put the video on a page on your website. Embed it on a page about the same topic, not just on YouTube.
• Include the full transcript. This is the part most businesses skip, and it’s the part that makes the video readable by AI.
• Add chapters. Timestamped chapters help people jump to the part they need, and Google can surface individual chapters as key moments.
• Add video schema. Structured data tells search engines exactly what the video is, how long it is, and what’s in it. Google’s video structured data documentation explains the details.

That combination turns a video into something every AI can read directly. This page is an example: the video is at the top, followed by the chapters, the description, and the full transcript, with video schema behind the scenes.

Section summary: Record in-depth answers to common questions, then publish each video on your site with a transcript, chapters, and video schema.

How to Get Recommended by ChatGPT

Here’s the full checklist, whether you’re trying to show up in ChatGPT, Gemini, Perplexity, or Google’s AI Overviews.

1. Answer buyer questions on your own site. Write pages that directly answer “who should I hire for this,” “how much does this cost,” and “what should I look for.” Put the answer near the top.
2. Get real reviews that describe the work. “Great company!” helps a little. “They rebuilt our Google Ads and our calls went up” helps a lot, because it ties you to a specific service and result.
3. Be consistent everywhere. Same business name, same services, same description on your website, Google Business Profile, LinkedIn, and directories.
4. Show your face. Put your name and face on your expertise with an about page, bylines, and video.
5. Publish your videos as text. Embed each video on your website with its transcript, chapters, and schema.
6. Stay on topic. Don’t produce a pile of random content that doesn’t matter. Go deep on the topics you want to be known for.

If you want the local-business version of this, we also wrote about why AI recommends your competitor instead of you, and the one fix to make this week.

Section summary: Answer buyer questions, earn specific reviews, stay consistent, show your face, publish video as text, and stay focused on your topic.

How to Test Whether AI Recommends You

Don’t guess whether AI recommends you. Test it.

1. Write down real customer questions. Use questions customers have actually asked you, or come up with five to ten you know they’d ask. Include “who’s the best person to help with…” style questions.
2. Ask each one in ChatGPT with search turned on. Ideally do it while logged out, so ChatGPT isn’t using your own history and memory to shape the answer.
3. Run the same questions regularly. AI answers vary from one run to the next, so one result isn’t the full picture. Run your set of questions often enough to see a pattern.
4. Check the sources it cites. This tells you which pages, reviews, or videos the AI is leaning on, whether they’re yours or a competitor’s.
5. Repeat monthly and track the change. Keep a simple log so you can see whether your work is moving the needle.

Section summary: Test real customer questions in ChatGPT, logged out, repeatedly, and track which sources it cites over time.

Ask Every New Lead One Question

Start doing this with every new lead, starting today. Ask one question:

“How did you find us?”

If AI was involved, ask them exactly what they typed. That prompt tells you which question you’re already answering well. It’s gold. In my case, “who would be the best person to help me set up a Google Ads campaign” told me that my personal brand and my Google Ads content were doing their job.

And if you’re winning that prompt, you’re winning. Now you know what to build more of.

Section summary: Ask every lead how they found you, and when AI was involved, capture the exact prompt they used.

FAQ

Can ChatGPT really send you clients?

Yes. When people ask ChatGPT who to hire, it often names specific businesses or people. In my case, someone in Los Angeles asked who would be the best person to help set up a Google Ads campaign, and ChatGPT recommended me by name, even though I’m based in North Carolina.

How does ChatGPT decide which business to recommend?

Nobody outside OpenAI knows the exact process. When search is turned on, ChatGPT looks up current web pages and combines them with what it already knows. It tends to favor businesses and people that show up clearly and consistently across multiple sources, like your website, reviews, directories, and content, especially when those sources directly answer the question being asked.

Can you pay to be recommended by ChatGPT?

Not through the normal recommendation process. The way to show up is to earn it by being easy to find, easy to verify, and clearly connected to the service people are asking about.

Do YouTube videos help you get recommended by AI?

They can. ChatGPT doesn’t watch videos, but it can read titles, descriptions, and transcripts. Google’s AI Overviews and Gemini cite YouTube frequently. Videos work best when they’re also embedded on your website with the transcript, chapters, and video schema.

Does being local matter for AI recommendations?

It depends on the question. If someone asks for a plumber near them, location matters a lot. If they ask for the best person for a specialized skill, AI may recommend someone far away, like it did with me.

How long does it take to get recommended by AI?

There’s no fixed timeline. In my case, it came after years of consistent work and a big increase in published content over the past few months. Businesses that already have strong reviews and helpful content tend to show up sooner.

Want AI to Recommend Your Business?

If your customers are asking ChatGPT who to hire in your industry, you want to be the answer. At PushLeads, we help small businesses get more leads without struggle and frustration, including building the kind of presence AI tools trust: content that answers buyer questions, reviews that describe your work, and video published as searchable text. Contact us, or run an audit of your own business on our website to see where you stand today.