How To Rank #1 On ChatGPT By... Gaming It?

Eric Siu and Neil Patel discuss how to rank high on ChatGPT by optimizing for AI discovery. A guest created a fake deodorant brand for $11 and hit ChatGPT's top results in a month by gaming the "AI shelf." They cover Answer Engine Optimization tactics: structured data, listicles, comparison articles, and high-authority citations.
Chapters
So, Neil, did you see the cafe in San Francisco that allows you to brand drinks?
No, I have not seen that.
So, check this out. So, there's a cafe here in San Francisco, right, where the startups can pay to brand drinks. So, ElevenLabs, you've heard of, right? So, you can do an Eleven Latte, right? So, iced double espresso shot, it's $5.08. So, $5.80 and then zero click, okay? You have Rippling, there's a Berry Ripped, okay? 21 grams of protein, 14— MongoDV Greens, right?
14.5. Okay.
Brexpresso, 7.5. And so I think this is cool. I honestly think that anything that hasn't become saturated yet, like this looks pretty saturated if I'm being honest, but anything that's not saturated where you can kind of stand out. So I have a friend that like, he likes to buy these airplane ads that like fly around the beach and all that. Or if you can, I think we had a mutual friend, he bought like, you know, when you land at the airport, there's like, when you walk the walkway, it's all like his company branding. And so sometimes you have to zig where other people are zagging. Like back in the day, podcast ads used to be unsaturated, right, where you can stand out. Or back in the day, influencer ads used to be not saturated, right? So if you can attack something where a lot of your people are hanging out, I think a cafe in San Francisco that's popular actually makes a lot of sense because there's a lot of foot traffic that goes through there and it's people that you want to be seeing your stuff. Yes.
And being the first is great. If 10 other startups do this or cafes do this in San Francisco, it's not going to be effective for the 3rd, 4th, 5th. 2nd maybe can get away with it. 3rd won't be as effective. But the first, whoever's the first to do it, it was great. It was like back in the day, do you remember Million Dollar Homepage where some kid was trying to generate a million bucks to pay for something?
Oh yeah.
I think college or whatever was selling pixel by pixel. Great model, but going and creating another Million Dollar Homepage is not going to be effective. It was already done.
Yeah. Marketing 101, Andrew Chen wrote this. It's the law of shitty click-throughs. Something might be great, the click-through rate might be astronomical in the beginning, just like email in the beginning, right? Then it just goes down. Back in the day, by the way, SMS, I think SMS is still pretty high, right, Neil?
It is extremely high.
It's still, because it's still very personal, right? But also like ChatGPT, click-through rates might be very high in the beginning. And Neil, do you want to see how this person ranked number 1 on ChatGPT?
Yeah.
Okay, so check this out. So this person made a deodorant brand that doesn't exist, and 3 weeks later ChatGPT ranked at number 1.
Okay, what did they do?
To strangers. I spent $11 on a domain. It took me 1 hour of work. I didn't do any weird Reddit stuff, no fake reviews. And here's why I did that. So I've written a tech newsletter for 5 years, and since June I've been mapping what I'm calling the AI shelf. So picture an actual shelf in a store. You ask ChatGPT what to buy, and there's a shortlist of brands it reaches for. That's this shelf. That's the AI shelf. I've run almost 9,000 queries at it. No one can see this shelf. It's opaque, but it sits in front of $6 trillion worth of global e-commerce, if not north of that. So I wanted to know, can you get on the shelf? And you can. So by day 21, I got my first hit. By day 30, ChatGPT with browsing on named Morrowind, and directly underneath it, this very real brand. One of them exists. One of them definitely does not. Now it only moved 2 of my 15 queries. Native still fully owns the category. I just found one little dusty shelf in the back of the store that was still open. Long tail is where the opportunity is. No surprise there. But beyond that, it reveals something about our relationship to shopping and AI that I think is going to have a huge impact. So you put something into Google and you get a list of links. You have some media literacy. You get to decide who to trust. What AI did with Merowyn is it handed the copy that I wrote back to me in its own voice. And that voice made it sound vetted. It made it sound like it had some authority. But of course it doesn't. And that's before agents own the whole loop from recommendation to purchase. So pretty soon I'm gonna be telling my agent to keep my cabinet stocked. So it's gonna pick, it's gonna pay. Trillions of dollars are gonna be moving through this door that currently no one can audit. It's very high stakes. Almost no one's watching it. I am. I test every month, so follow along. I made a—
So all she did was just use text on her own website and it started crawling it, and then she started doing it.
It sounds like she uses her site as like a promotional, uh, vehicle. And, um, but I don't know what she meant by like this, um, AI shelf or whatever. So I don't quite understand that, but let me tell you what I would do, Neil. I would just take this Instagram link, I would dump it into my Codex desktop, and I would say, make this into a skill. And I don't even need to do anything. And I'll just send it to my team and say, hey, take it from 50%, get it to 80%, 90%, and then we're good to go.
Yeah, I do think the copy on your website has a big impact because no matter what people write about you on the internet, if your website doesn't show that you do it, It is hard to end up getting placed high in ChatGPT or getting placed at all. The second thing is listicles. We all know about that. The third thing is comparisons. When you do comparison articles of you versus other premium brands that they're recommending, that does well. The fourth thing, we've been doing a lot of this at my agency, NP Digital. We look at all the sources they're citing and we try to get those sources to just include us. See a lot of people like, I need listicles of the 10 best You know, toaster ovens or laptops. I'm like, you do know if you type in all the queries, they're already pulling from certain sources. What are those sources? And why not try to convince those exact sources to also list you? That is an easier way to get cited than anything else. From everything we've tested, if you can just convince existing sources to include you, it beats everything else out by far.
And guys, I'm sorry to oversimplify this, but to me, AEO is, there's, you have more digital PR, you have more structuring that needs to happen, and it's just more of SEO. So people don't want to hear it. And it's still, I still find it hilarious, Neil, when people come to work with us and they're like, we got SEO down, we just need help on the AEO. And that actually means they don't have SEO down.
Uh-huh.
Right?
Yeah. It's just like, oh yeah, what's your revenue through SEO? Either we don't know or it's declining or slightly up or flat. And it's just like, let's see what you're doing. Oh, this is what you're doing wrong. And this is also why you're not doing well on LLMs. Yes, there are nuances and differences, but I would say 80+% is very similar.
Yep.
So let me share this one over here. So Neil, do you see this graph?
Uh, KIMI downloads and users skyrocket. Yeah, I believe this. I believe it's going to just continually hockey stick up for a while. And we haven't even really seen the popularity unless something new ends up coming out that everyone ends up loving.
I'll shed some context on this one just for people that can't see the screen. So what I have pulled up is KIMI-K3, which is one of the best open source weights out there. So their downloads and users skyrocket. And so you can see here the number of downloads shot up from— this is daily, by the way, so it shot up to 1.8 million. downloads. And I think Daily Active Users shot up from about, let's say, uh, you know, 75K, 70K or so to about 1.4 million. Um, and then it started to kind of, you know, peter off a little bit, right? So my point of even showing you this chart over here, um, is when you think about the marketing behind Kimi K3, it's not like they did some crazy advertising campaign. Same Starlink, like they, they run ads, but it's not like a lot, right? But to me, Real quick, if you want to acquire customers faster and more efficiently this year with the latest strategies and tactics, then check out singlegrain.com. That is my ad agency. Again, www.singlegrain.com. Check it out. And if it seems like a fit, we'll get in touch and help you with a free marketing plan. Neil, I like— I've heard about people avoiding certain routes just so they can fly with Starlink on the plane. Okay. And then because the product is so good, right? Then you look at Kimi K3, because the product is so good, people talk about it being, you know, close to Fable levels, Fable 5 levels, which I think is is a little bit of exaggeration, but it's close, right? The point is these things are getting so good, people are like, there's a clear problem to be solved, whether it's Starlink or you look at Kimi. And on the flip side, like I mentioned earlier, when you have such good marketing around something like a— OpenClaw had all the mindshare in the world, right? But they couldn't execute anymore. They became unreliable. And so Hermeus started to take their mindshare, right? And then now you have other harnesses doing similar things. You have GrokBot that's kind of similar. And so I think the fundamental piece of it at the end of the day is this. If you have an amazing product or service, great, you can polish it. But if you have a poo-poo product service, if you put gold, you polish a turd, it's still a turd at the end of the day.
Yeah. With what Eric was just sharing on the screen right now, it's another marketable moment. a16z does this. We've been doing it for a very long time and we do it in a different way than a16z. We take a lot of our own proprietary data, our own research, return on surveying and create charts from it. Eric was showing a chart from a16z. They took 100% of the data from Sensor Tower, I believe it was, ran it to a16z, cited the source, which is fine, and it's getting a16z name out there as a venture fund, even though they're already well known and they probably don't need their name out there anymore. But the point I'm getting at is you can create marketable moments for your business, whether you have proprietary data or not. A16z is a prime example of this. I've seen a lot of the other charts where they're just taking data from like places like SimilarWeb. There's a lot of things that you can do like this to go and create brand recognition for yourself by just aggregating the data, putting it in a pretty format, and then sharing it with the world. You may not get as many views as you want, but if you're creating the right type of charts on the right type of data, it may really focus on your ICP and that can drive a lot of revenue in the long run.
Yeah, yeah, we should spend more time on marketable moments because this is something I'm pushing the team on a lot more now. And we've talked about last week that every, every person is becoming part-time engineer and part-time marketer. Okay.
Yeah.
So I told the engineers, I said, hey guys, like, look, you're all going to become part-time marketers now. I'm going to show you how to do it. And so whenever we launch a new ClickFlow feature, for example, I have this shared on the screen over here. You can see that we have a little video kind of outline. This is a Hyperframes video, and then people can spin up these videos very quickly. And you can see, oh, this is what this new feature does over here, right? And then actually this one got 26K views. It actually got 40K views on X because I did like a quote tweet. And so there's that, right? But who did I get this idea from? Well, you see a lot of it on X already. So Matt Van Horn, he's always launching new stuff. And so today he launches Agent Cookie for GrokBot, right? And he always has good music tied to it. So let me play some of the music over here with this video.
Real pretty, them flips impossible, man. Like they jumping like the laws aren't logical. Up, down, breakneck, alien vibe.
Okay, so he really likes rap, right? And so he'd always send me like a new video and he like texts me like, hey, here's— what do you think about this new rap song? Right? And so it can— it allows you to inject your own personality too. But here's the thing. So Neil's talking about charts and things like that. I think charts are very good. I think videos are good too. But beyond that, I think it also depends on what you're publishing too. And so we have the Skills Dojo where we have all these skills. Well, the easy thing is just keep promoting amazing skills that we find on X and then add them to the Skills Dojo and give them credit. And we can keep promoting that over and over and over. That never stops. I can make short-form content for that too, right? And so there's that. Our API gateway where we allow you to connect different tools, different connectors, have different permissioning, security, governance, all that stuff. That is stuff that I can promote as well. And so that makes it more exciting for the people creating it, and it makes it easier for us to continually have this marketing machine. And this whole concept of marketable moments, I actually got it from our mutual friend Patrick Campbell, where when he was running, I think, ProfitWell back in the day, it was Every month, right? This is like years and years ago, right? There's no AI, right? And years and years ago is like every month we need to talk about a new thing. We need to keep our audience excited. We need to, you know, get them, get them on top of this. And I think now one of the things is as a marketer, I can launch a lot more of these, um, on demand because I'm a lot less resource constrained, um, with the ability to do these things.
Yeah, dude, we all can just do so much more that we couldn't do before. And there's no more excuses for just not getting stuff done.
I got a question for you because we're talking about getting stuff done. This can be the last topic here, Neil. Do you think there are net more jobs in Philippines now or less jobs based on how the last few months have been going?
More jobs. I think by far there's more jobs because people want cheap labor to go and do all this stuff for them.
Yep. So, so I was trying to throw you off by throwing in something random at the end of it so you couldn't guess. I wasn't trying to lead the witness, but let me show you this. So check this out. So what I have on the screen over here, this is from Truong Phan. And so he's like, look, since the launch of ChatGPT, the Philippines offshore industry, so 8% of the country's GDP, I didn't know it was that high, has actually grown much bigger. So employment in IT and business outsourcing is up 20% to 1.9 million workers, and industry revenue is up 30% to $42 billion. So AI is helping offshore workers. If you look at this graph over here, it's a Back office bonanza. Okay, so Philippines IT and business services. You look at employees, 2016, it was a little over a million, um, and then now it's, oh, it's close to 2 million or so. Revenue back in 2016, uh, this looks like a little over $20 billion. It's called like $23 billion or so, and now it's, it's $42 billion. Okay, so what's happening now is AI is helping offshore workers in the Philippines do new, more complex jobs. They're picking up work training AI models or supervising AI agents. Hospitals in America are increasingly outsourcing the checking of insurance eligibility and filing of medical records to Filipinos packing AI tools. So, more high-value work such as accounting or engineering is going offshore in part because AI is a leveler. You can now teach someone something complex. You can teach someone complex stuff more quickly and it can still get done more cheaply than America or Europe. And so, people might think, oh man, Uh, all these, uh, BPOs, all these offshoring, like they're gonna go out of business because we don't need them anymore, right? But nope, more. And we talked about this for a while, more.
I, I think it's all going to be more for a while, you know. And, uh, I even watched a lot of the news channels and like, yeah, you may not need as many, as much memory. Well, what happens when there's robots and all this kind of stuff? Like the memory usage will be more. I'm not saying memory, the profit for a lot of these companies will be the same.
I'm just saying they'll need Do you think you'll ever fall in love with one of your robots?
If it can clean 24/7 and have the house spotless, yes, because I'm so OCD.
Uh, all right, well, I think that's a good place to end. Uh, guys, we will catch you next week, and, uh, yeah, have a good weekend.










