Companies fail with AI because of this, podcast mention drives $29M in revenue, Brutal new SEO stats

4th. You know, Neil, it's, it's one day before I'm no longer in my 30s. So, uh, yeah, going into my big 4-0. Yeah.

Happy early birthday.

No, no, no, it's okay. I'm not there yet. So you were there, you were there last year.

Well, happy early birthday, because for me, I was gonna say I have it in my calendar. I have it in my calendar, but my calendar is not giving me, you know, the notification that it's your birthday yet. But my calendar is in PST even though I'm in Indonesia. So that's why I'm not getting the notification.

Yeah. So Neil's recording, it's 4:00 AM his time right now in Jakarta. And then I'm still in LA right now. So we're gonna go ahead and get started. We're coming to you live right now. And Neil, I think maybe this one's fun to start with. So I saw this, I don't know if you saw this, but Harry Stebbings did an interview with this one guy. I think he runs a company called Corgi Insurance and I haven't heard of it before, but apparently they're doing really well. And I think this guy's in his early 20s. And so one thing he said, Neil, was if you're not working 7 days a week, you will lose.

Okay?

And I think we have our thoughts around this. Obviously, we're in our 20s before, but I just wanna call this out. So he says, look, this guy from Corgi Insurance, if you're not working 7 days a week, you are going to lose. Corgi Insurance is the most intense workplace culture in startups. The company works 7 days a week. The founder lives and sleeps in the office. He built a cafe in the office because there was no local cafe that was open 24/7. 2/3 of the first 30 team members have the Corgi logo as a tattoo. Okay. And then today—

Wow.

He went behind the scenes. I don't know how he got that happen, but he went behind the scenes. So they scaled to a $2.6 billion valuation in just 2 years. And so here's, we'll call out some more things here and then we'll react to this. Work trials repel the mediocre. So Corgis force candidates into mock work trials over the weekend. To test them out to see if they're intense. Lead from the front lines. If you can't demand 7 days while working on a yacht, so he sleeps there 3 to 4 hours. He only sleeps 3 to 4 hours a night. And then culture only means one thing, winning. And then he talks about building something world historic requires radical sacrifice. When asked if he'd rather build a trillion-dollar company and die at 50 or fail and live to 80, the answer was easy. I would rather measure my lifespan in victories. I think you like that, Neil. So anyway, all that to say is I think we certainly had a way of thinking like this in our early 20s or so, but we've learned with age, at least I'm speaking for myself right now, that it's also a marathon, not a sprint. So you go first and then I can react.

Well, first off, I love the— people are probably going to hate me for this, but I love how we chose the winning and dying at 50, I personally wouldn't want to live to 80 and have my life be tons of failures, right? The successes— and I have failed, I know most entrepreneurs have failed one time or another. Even if your company was a success, there was times during that company that you probably had many failures. Um, but the successes, the reaching some of your goals and then creating new ones, that keeps me going. And it makes life really enjoyable for me. So I would choose dying earlier and having, you know, hitting some sort of milestone. I'm not saying I have to need to create a trillion-dollar company, but I wouldn't want to live till 80 and just always failing in life.

Let me ask you a question. So if you, if you went back to your 18-year-old self right now, and then he looked at you right now, um, would he consider you a massive success or failure?

Because you said 18, it would be a success.

Okay. 25-year-old?

Okay. 30-year-old failure.

Well, by normal measures, and we're not measuring normally right now, but I think if you look at the reason why I asked the 18-year-old question is I would say that it just depends on the beauty's in the eye of the beholder, right? And so for you, because you're always stretching, stretching, stretching, the goals always get bigger and bigger over time. So I just look at it this way. I enjoy playing the game of business and I just want to play it to see how far I can go. And if it becomes really big, then great. But I also know that Neil, you like a nice life. I don't think you would be okay living in the office and sleeping 3 to 4 hours a day, would you?

No, I really do enjoy a nice life. I could not work like this corgi guy. Does. When I was 18, I could. I'm assuming he's older than 18, but I could not do that. Like even Elon, they say, like, I think people were fundraising to put a bed in his office because he was sleeping on a couch. I can't do that. Yeah.

Yeah. So I'm the same, right? So if we go down the list over here, so I don't necessarily believe that. Well, actually, no, I don't know, Neil, if you're not working 7 days a week, you're going to lose. I think if you're the entrepreneur and if you're not working 7 days a week in the early days, yes, I do think you're going to lose. That I agree with. But I don't think that applies to necessarily like all your employees. That's what I would say. Yeah.

And even right now, I think both you and I work 7 days a week still. Do we work as hard on weekends? Sometimes yes, sometimes no. I think it really varies on what we're doing, but sometimes we're working on some projects and we just want to crank it out. But I would say typically we work the hardest Monday through Thursday. I would say my Fridays aren't as hard as Monday or Tuesday or Wednesday or Thursday, and I would say my Saturdays and Sundays are definitely more chill than even Fridays. Um, and yeah, it also is what do you want from life? Like, I started in marketing because I accidentally landed on it. I wasn't intentionally trying to be in marketing. I was just trying to make money. And, you know, I had my own website. I I did marketing for it. The website got a lot of traffic. I didn't make any money. I was taking a speech class while I was in high school, but it was a nighttime college class. And I gave a speech on how Google's algorithm works and how to get traffic. It was a random topic that no one in the class really cared for except one person. And that one person worked at a company that was looking to hire someone like me. They were paying me over $1,000 an hour for consulting. It was a flat fee, but the amount of hours that it took, it was over $1,000 an hour. I was ecstatic. And then from there they gave me a $5,000 a month contract and I was 16. So I was like, huh, I can end up having a nice life. But the problem with the nice life is once you start buying things like nice homes or nice cars, it's never ending. And you continually want nicer cars, nicer homes. After a while, I don't know how much nicer of a home you can end up getting, at least for me. The cars, yes, I can definitely get nicer cars, but I don't care for it. The home, the only reason I spend— and I know people say it's a really bad investment if people knew what I spent, you know, on construction of my home, building it, etc.

You've said it in public before.

Yeah, so call it— I'm in it, you know, 30, 31, 32. It's somewhere around there because of construction costs. I don't know the exact number, but I definitely know I spent more than 30. I definitely know I spent less than 32. Uh, and that's without finishing the backyard. So let's call it, it'll end up being $33 or $34 million. From a money standpoint, it is extreme waste. And my house isn't big, or it depends what you consider big. But, you know, people consider mansions, I think they're defined at 10,000 square feet or more. I'm living on 7,000 square feet. Um, the biggest cost for my house was actually the land. Outside of the land is the structure and whatever faucet you pick or furniture or whatever it may be. But I look at it as the house makes life super convenient, especially if you have a family. It just makes your life so much easier and more relaxing and enjoyable. Um, so I like that because for me it makes me work in a happier place because I work from home and it makes me more productive. But other than the house, I don't think I would spend money on anything other thing that would be that crazy in expense. Like clothes, I think they're a waste. I've spent my money on that. Cars, I really think they're a waste. I have a Tesla Model Y, as you know. The seats suck, but the click, type in your address and click and it just drives you, amazing feature.

You know what's funny? Neil keeps talking about how the Tesla Model Y, the seats suck, but I'm just like, well, you have the money to upgrade the seats, so why don't you just upgrade the seats?

It becomes a safety hazard. So when you have a car, it's tested. When you start changing parts, or, you know, when people like lower a car or raise a car, it messes up the sensors for when someone hits a bumper.

You know, my back hurts too, right? So I just put one of these things in the back too, so it feels really good, right? So why can't you just have some of these modifications for your back and just make it better? But anyway, I— we digress right now. So I want to call out some other things here, and then we can move on from this, this this topic. But yes, I do think you have to, it's up to you. It's not what Neil's saying, what this corgi guy's saying, or what I'm saying as well. You have to decide what's important to you in your life. And I applaud this guy. I think Neil applauds this guy too. This guy's name's Nico. If that's what you want in your life right now and you're going for a trillion-dollar outcome, more power to you. We support you for it, right? So let's go through it. Here's some things that I do like about what he's saying. So check this out. I do like the fact that he built a cafe in the office because there's no local cafe that was open 24/7. Because here's the thing, Neil, if you have a 24/7 office, guess what? It's in San Francisco. Who's going to be doing deals at the office? You can bet the best VCs, the best people, the best investors, the best founders, they're going to be in that office. And also you're going to have some of the most cracked out talent. I think this is a nice way to actually get talent. So I think that's really smart. The second piece over here, and I don't know how much you guys do this, but We used to do this quite a bit where work trials repel the mediocre. So Corgi forces candidates into mock work trials over the weekend. If seeing a full office on a Saturday scares them, they don't belong. True intensity acts as a filter, a natural filter to attract killers and repel clock watchers. So here's what I'll say about that. I like paid work trials. I'm not saying I want people to come into an office over the weekend, but I think it's, even though it's expensive to do a work trial, it's more expensive to hire a bad employee. And so I, I, this just reminds me that I need to get back to it. Like, are you guys still doing paid work trials? You guys are, right?

Yeah, we've always been doing them. We think it's one of the best ways to figure out if someone's good or not. Paid work trials don't work for all positions though. So if you're hired, let's say in marketing, you're hiring someone to do SEO, a paid work trial isn't too hard to figure out if someone's good or not. If you're looking for a managerial role, let's say a managing director of a certain region or division, You can get a sense if you like the person over a paid work trial over the weekend, but it's still not going to give you enough insights if the person's going to work or not.

Yeah. I think managerial is more nebulous and bubbly versus do you know how to actually do the work or not, whether it's paid media, SEO, AEO, whatever it is exactly. And so lifespan versus victory, I think that's interesting. I do think we should lead from the front lines. Look, even though Neil's been working for a very long time, I think 24 years in business, I think I've been in the business for about 13 years or so, we do try to lead by example. And even though Neil likes to say, oh, I don't have much going on or whatever, no, but Neil still works really hard right now and he does try to lead by example. Same thing for me. Any other thing here that sticks out? Oh, reject the comfort of quiet quitting. If you're operating in a hypergrowth environment and your days off happen to be Saturday and Sunday every week, You're quiet quitting. To win, you must deliberately bypass the off-ramps of personal comfort, low volatility. Okay, I don't agree with that part, but to each his own. Good.

There's a lot of companies that were built where most of the employees did not work Saturday and Sunday. I'm not saying working Saturday and Sunday is terrible. I've seen it work extremely well. You and I do it. But I do believe some people, not all, some people need a recharge.

Yep, I agree with that. I mean, I would say the vast majority of people cannot sleep 4 hours. That's actually, I don't know if you know this, Neil, because I'm a sleep nerd, but there is a group of people that can only sleep 4 hours and it's like a specialized gene where it basically, they can just keep going. So I want to show you something over here because we're talking about employees right now. So check this out, Neil. Don't worry, Neil, I'll carry this because it's 4:00 AM over there. So AI is creating a new class of super IC. So an IC is an individual contributor. So the era of 10x engineers, PMs, product managers, designers, AEs is over. We are now in the age of the 50x employee, okay? So it's essentially a reset. Anyone can unlock massive career benefits by leaning into AI. So check this out. This is from Cursor. So when you look at the output gap is widening, we see P90 developers. So I think that's percentile. So 90th percentile developers pulling farther away from median developers in absolute lines of code added per week. So P50 means you're in the percentile of 50. P90 means your top 90 percentile. So if you look at this over here, you can see over time, the P90s are getting higher and higher over time versus the P50. So it's never been more expensive to be average. And then sure, we're looking at developers right now, but I think you can apply this to really anything across the board. And I know lines of code is just one metric, but I mean, come on.

Well, when I put it in a marketing perspective, because I've seen this a lot where everyone talks about The 10x employee, the 50x employee, the 100x employee. I have not seen A players become 100x, 50x, or even anywhere near 10x better from leveraging AI, at least in the marketing realm yet. I'm not saying it doesn't make 'em better. I just have not seen the magnitudes of an A player becoming 50x or anywhere near there better. Forget even 50x, like 5x. I haven't seen an A player even go 5x better. From just leveraging AI.

So I'll give you a response to that in a second after this chart over here. So P99 devs, so these are basically top 1%, produce 46x more lines than median dev. I predict this comes to every function. In my opinion, many top performers will be constrained by the old system, couldn't design code, lacked organization. AI gives them these powers. That's exactly what I wanted to cover. So Neil, I would just say that, for example, we have someone on the sales team right now that just came from paid media, similar to your guy, came from paid media, not on the sales team. He would've had to wait weeks to get a landing page up because he wants to run ChatGPT ads. He basically got his mockups, the ad mockups ready. He got the landing page up and he deployed the landing page as well. That would've taken weeks before. I think what AI gives you and me, Neil, is powers that we didn't have before. where we wouldn't have to wait a long time to deploy certain things. And I think in that respect, it makes me stronger because every job is just a list of responsibilities. So it enables me to do more with less, right? And so I do think it's going to come to fruition at some point. I think you will see it, Neil. I just think it's happening case by case right now. It's one area. Coding's very apparent right now.

Yeah, coding is really apparent, but outside of coding, specifically marketing, I just haven't seen— I've seen efficiencies. I haven't seen efficiencies in multiples. I agree with you, you can get landing pages up faster or do keyword research faster or figure out errors on a site faster. But when you look at what someone's spending 8 hours a day on, for most roles in marketing, AI is not making them 5 times more efficient or whatnot, right? There's certain tasks that it is making them more efficient in. But there's still a lot of stuff where AI isn't helping yet and time will tell what happens there. But a great example of this is I had a team member come to me the other day and he's like, Neil, I want to do an experiment where we start cranking out AI content on your social profiles and we're going to do text-based. So I said, all right, are you talking about short form or articles? They're like mid-form. And I said, if you're going to do it, I recommend articles. Because a lot of these social networks are giving favoritism like X and LinkedIn if you want to start posting articles because not enough people do. They're like, no, I want to start on Midform. I'm like, all right, go for it. You do your Midform experiment. So they start cranking out Midform content. I don't know what the length is, but it's not 200-300 words. It's not 1,000-2,000 words. It's somewhere in between. And they're cranking it out using AI. And I was like, all right, we can test a few of them. So we tested out for a few days. And I told them right away it's not going to work. They're like, why? Because they showed me the content before it was posted and I tweaked it a little bit here or there, but I didn't want to tweak it too much because I wanted them to see the actual results from quote unquote AI content. And it got posted. It didn't get tons more engagement than a normal post from a view perspective or anything like that. And, uh, we reconvened and they said, hey, can you give me another 2, 3 days? I'm gonna make some weeks. I said, no problem, I'll give you a week. So I'm like, but just keep learning and updating me with your learnings. And they asked, why don't you think it worked? And I said, the problem with AI content is if you look at what's being written, there's nothing new. You're regurgitating the same information. I'm like, that's the first thing, because the way they're doing it is they're scraping to see what's worked for other people and then coming up with their own twist of that version. And I'm like, that doesn't work too well because once it's already done well on social media and a lot of people have seen it, even if you do your own version of it, it's not going to be as good or it's not going to be as exciting just because people already have seen similar stuff before. The second thing is when you look at AI-generated content, even if you have it help you fine-tune the hooks and make it a little bit more punchy or use stats and data, it still comes across as AI-written content no matter what you do if, and this is the big if, if you've been creating content for a long time. So I've been publishing content on the internet for more than 20 years. People understand my tone of voice, my writing style, and yes, you can get AI to train to write like you, but it's still not the same if you're decent at writing, or I would say if you're good at writing. And I wouldn't say that's the biggest reason. I think number one is the biggest reason. The writing style and the way it's doing hooks is still not perfect. And it's not going to learn in a week, even though they uploaded, no joke, years of my content. I think they uploaded over 1,000 articles to train it. It's just still not perfect. And I think writing is part of it, but the bigger reason is It's the topics that it's picking and it's not offering something new. Well, versus if I were to create an article on LinkedIn or X, I would pick, or even a post, I would try to talk about something that's new.

So I'll kind of relate to this and I have a cool, I have another cool chart I'll show you. So I was talking to Alex Lieberman again this week and so he runs a company called TenX, which is an AI transformation company. And basically I was like, hey, so you have this content machine skill, tell me how it works. And he's like, I'm like, okay, well tell me what results you're seeing from this content machine skill. So basically it drives millions of impressions on X every month and also on LinkedIn. I was like, okay, well how's it doing for you in terms of driving pipeline? He's like, it helps me drive $100 million plus leads from LinkedIn and that's where we get all of our pipeline from. I'm like, okay. Well, tell me more about this Content Machine skill. So basically, long story short, he gives me access to the skill and I've implemented it into my Hermes and I tried it again yesterday for the first time when I was bored. So after I got off all of my meetings for the day, let me share my screen with you. This is how it works. And then you guys can take this for yourselves and go from there. So yesterday I have, you can see it says use the Content Machine to quiz me in a thread. And this was at 5:32 PM yesterday. And when I was asking Alex, I was like, hey, this thing interviews you, but that's really annoying. How long does it take to interview you? And he's like, 20 minutes. And so the way Alex uses this is it interviews him and it comes up with cool angles and then it'll write the stuff out and then he can choose what to edit. But he just wants to be at the beginning part, let the middle part AI can handle, and then he handles the end, the editing. So he's like, so this thing starts to interview me and I'm not going to go through all of it, but it's like, hey, what happened that made this a great idea worth capturing today? And so then I just start talking to it back and forth. It just keeps asking me questions. There's 10 questions. And ultimately right before this call, Neil, I talked to the social media manager from my team and he's like, whoa, this idea is really good. But to your point, Neil, it's written in such an AI way that we're like, it's such, it's so part of my language. It's so crappy the way it was written, but we know the seedling of the idea is really good. And he's like, I was like, obviously you and me, Neil, we live with ourselves all the time because we're in our bodies. But for him, it's super helpful just to get it out. So now I can run this whenever I'm on the go and then just give cool ideas in there. And then basically I'm just stealing Alex Lieberman's content machine skill, but that makes it a lot easier because Neil, you and I, I think I won't speak for you, but for me at least, because I'm so running around all the time doing random things, it's hard to just sit down sometimes and actually get this idea out. And usually it's the idea that goes really far. Also written in a good way too. That's what I would say.

So this model on how Alex Lieberman is creating content is actually very similar to how I do my social content. So I get on the phone with AJ, I bring up all my ideas and what I want to talk about. He then does the same with my team because they may have some unique angles. Then AJ goes and creates a script. He mainly uses AI to help him create the script. He then goes and reads the script and modifies and tweaks it and whatever. Then he'll send it to me and then I'll go and adjust it. And then boom, I will go sit down and we'll crank out like 4 or 5 of them in one sitting and just click the record button and there's a teleprompter in front of me. That model's worked out really well. The other way I found AI to work really well for me is I have a ton of ideas. I then go take the ideas, I put it into AI, and then I have it help me refine the ideas or come up with other unique angles that I'm not thinking about. So in many cases I may have idea for SEO or creating content for specifically Google, and maybe that same framework would work for social media. So then it'll respin it and be like, hey, what about using the same framework for social media? Then I would go in and be like, well, social media is a little bit different because more of it's video versus text, and I can modify A, B, and C and make it relevant for it. But I found those 2 approaches to work really well for me. I haven't found AI from soup to nuts to work really well and get the engagement and likes that I want unless— and this is the big caveat— unless I have AI produce me content on something that's really hot. So if Google Marketing Live just happened or Google, uh, uh, I/O just happened, right, their main event, and we start pulling in main things from I/O And just start creating content around it, it tends to do well because you're riding a wave of something that's trending, right? Similar to when you started creating Twitter articles on OpenClaw, and I don't know if you do it on Hermes or whatever else you're writing it on. If you're creating it on what's hot right now that everyone on that platform's talking about, it typically does well, even if it's AI written and people can tell.

And that's true. Let me just show you this, Neil, and then we can continue on here. This is my X analytics in the last 12 months or so. So you can see nothing, nothing, nothing. And then obviously December happens and then I'm like, oh, okay, we're moving into a new world. Then January, I'm going to read this off. Okay, so December, 183,000 impressions. January, 267,000. February, 831,000 impressions. March, 1.4 million. And then April, 1.2 million. And then May comes, 2.3 million. And then we're in June right now. June just started. So all that to say, I think this will continue to compound. But Neil, to your point, a lot of the articles that are written right now are using skills that we have and are talking about AI transformation. Now we're talking about the single brain stuff, right? X is perfect for it because it likes that content.

Yes.

And so that's what I would say. And so anyway, all that to say is there's caveats to it, but I think the best way if you want to produce the best content is having something interview you. And the cool thing is, Neil, the skill that I made, the thing that interviews me, I can keep adding to it. I can keep adding it. Hey, here's my voice skill over here. Hey, here's how I would go about creating LinkedIn artifacts, for example. There's all these things you can keep adding to it. Whereas I think if you're working with a person, like a human, it's still good, but I want to keep adding to it over time because I want to keep tweaking. I want to keep adding all these upgrades.

Dude, I'm with you. But the key right there is, and you nailed it, you're riding the wave, right? You're just following the trends that are really hot right now. That model works. People just tend not to do that. If you end up just riding the wave on what's hot, even if your content is mediocre, and I'm not saying Eric's content's mediocre, but if you were to just do it or if someone listening was to do it and their content was mediocre, it would still probably do well. Like the key is just, it's better to be on it and early on a hot topic than it is to be late on a hot topic and have your writing or your post being perfect. I'd rather have a messy, not as well done post, but being early on a hot topic, and that tends to do much better in the long run.

Dude, so check this out because we're on the topic of AI right now. So Aaron Levy responds to this, but this is a graph from the Kobisi letter. Kobisi. Okay. So what if AI is actually creating more jobs than it is replacing? So the latest JOLTS data showed that US job openings surged by a massive 731,000 in April. That's the green line, right? The job openings. So when you look at 2008, this is when the GFC was happening over here. You can see the number of job openings really dropped a lot.

What's GFC?

Great Financial Crisis.

Oh, okay.

2008. Yeah.

Yeah, the financial boom, the mortgage.

And then what happened is job openings percentage changed from a year ago. Like it starts going up, but the number of actual job openings, green is what you should be paying attention to. So this, you know, Let's call it the great financial crisis, really dropped job openings down a lot. And then you can just see it starting to rise, rise, rise, rise, rise. And then 2020, okay, the pandemic hits and there's a sharp V, and then they call this a V-shaped recovery. So boom, the stimulus comes in and the number of jobs shoots up. But then what happens is, oh, interest rates start going up, right? And then there's a lot more that goes into that. The war happens, right?

All these things.

So jobs start going down again. But guess what? It's starting to come up because everyone was like, oh, AI's gonna start to really destroy the job markets. I think there's gonna be job displacement, but I think net-net there's gonna be a lot more jobs available. So remember we've been talking about this recently, AI, at least my stance on this is that AI's actually gonna create more jobs long-term than destroy jobs.

So we see it in marketing too. We believe AI is actually gonna create more jobs than it displaces. It doesn't mean AI won't get rid of jobs, but it'll create new jobs, new categories. And we think the new categories and the new jobs it creates outweighs the jobs that it displaces, at least in marketing.

Here's how I see it, Neil. I think if we go back 100 years, right? Or 150 years, you and I might be farmers, right? Because that was the most common job and we'd probably fail at that, right? If I'll speak for myself, I don't think I'll speak for you too. I don't think you'd be a good farmer either, but like, no. My point is you have what, it was like 92 or 98% of the population were farmers. And so these are jobs nobody wanted to do. Okay. People are like, oh, what's going to happen to all like the Uber drivers, for example, the taxi drivers? I think what's going to happen is one, I don't really think they really enjoy their jobs per se. And that's just my hunch. My hunch is they're going to be able to do something that they really want to do and be able to take care of their friends and family a lot better. That's what I think. Yeah.

And speaking of this jobs data, what Aaron Levy here, he's a Box CEO, is talking about is AI has disrupted many categories. And one of the most obvious ones that is disrupted is engineering. And everyone's like, oh, you won't need as many engineers. But if you look at the data that he's sharing, more people are running engineering products or software products. So it's causing people to actually want to hire more engineers because they're able to do a lot more with less. And we're seeing this in marketing. Eric knows this for probably 8 to 10 years now, I've been pitching people in marketing how they should release free tools because free tools is a great way to get brand mentions, backlinks. Although back in the day, I wasn't talking about brand mentions, more so talking about backlinks, getting more branded searches. And it's one of the most effective strategies. And I think in many categories, it's actually more effective than just paying for advertising. That's why I like the concept of just giving away tools for free. And we've done this concept with Ubersuggest. I tested it out with AnswerThePublic as well. And we have quite a few other tools on our websites. Some are just simple like A/B testing calculators and the list goes on and on. What we found is when I've given talks on how to build a big company, giving away free tools, collecting leads from those free tools and then selling them on something else. So for us at NP Digital, we give away a lot of free tools. Collect leads, sell those leads on marketing services. Yes, we do have paid versions of our tools, but some of our tools are purely for free. What we found is when I gave that pitch or speech at events and shared the data, very few people did it. Want to guess the reason why very few people did it, Eric?

Because it takes work.

Takes work, and more importantly, that work costs a lot of money. I'm finding so many people to hit me up asking me questions about how to convert tool traffic into revenue. Like, no joke, dude, I'm getting 2 to 3 inquiries a day. I mean, not a day, a week about it. And when I talk to those individuals, the reason they're willing to do it now is because AI makes it way more cost effective. So they still hire an engineer that they may have not had before, but instead of having to hire 5 or 10 engineers and pay all these expenses, They're hiring one engineer or a part-time engineer and using AI and then getting something that will get them initial traction. Now, most of them still aren't looking to go full-blown. They're just looking to get something out there and see what traction it gets and then spend more over time. But you're seeing more companies, or at least more marketing departments willing to invest in engineering products to get their brand out there, to get the backlinks, to get the brand mentions. And they wouldn't have been willing to do this 2, 3 years ago because the cost was ridiculously high compared to now. So you're seeing companies just willing to do a lot more. And another example of this is Legion, the protein powder that I take and the protein bars you've tried.

Uh-huh.

So I was talking to their CEO and he was telling me how Matt— Matt's a mutual friend of ours. I'm the one who actually introduced him to him and he hired him. He's been there probably for like 8 to 10 years now. Matt runs his engineering. He's like, dude, Our engineering guys get way more done. And he's like, people would say that we don't need any more engineers, but he's like, because it's so much more cost efficient, we've actually increased the number of projects and the output that we've been, you know, we've increased the output as well. We're now, because it's so affordable, we're actually thinking about hiring a few more engineers and just exploding our product pipeline and just doing way more because it's so much more cost effective and the economics make more sense now.

You know what's funny? Let me just go on a little tangent here and we'll come back to this. We hired a software engineer from Oracle, okay? Based in India, okay? Because of that point, Neil, I think because this guy's so good, they did everything they could to retain this person. Which is interesting to me because people are like, oh, we need to keep the really good engineers. And I think you're starting to see the counteroffers come out now, right?

Yes.

So I'll just leave that one at that because we ended up coming to another arrangement with this guy. But I think we're going to need to hire more because of this point over here. So again, job openings soar by 731,000, which kind of, it changes the narrative that AI is actually creating new jobs, right? So AI is going to cause companies to hire more in sales because agents can let them process more leads and do more customer research. AI will cause an explosion of new marketing roles Because of how much more efficient it is to launch campaigns and target. The list goes on. So the thing is, I'm noticing in these single brain conversations that we're having for just this AI revenue agent implementation, everyone's like, can you do this? Can you do— it's always every single call. It's not like them trying to poke holes in how your campaign's going or whatever, or talking about the results. It's like, how about this? How about this? How about this? How about— it's constantly more, more, more, more, more. And so going back to the initial topic that we had where the guy wants everything, he wants to build a trillion-dollar company, I think that's going to be much more, it's going to be hard still, it's going to be much more accessible is my point here. Yeah.

And dude, I don't know what chunk of your clients is B2B, but I'm assuming a good portion.

Yeah.

Yeah. Same with us. I would say it's probably our biggest portion is B2B. Is that for you as well?

Yep.

So what we're seeing in marketing for B2B is A lot of our clients are software companies in B2B. We have security in other sectors as well, but let's just stick with software for now. Software companies would produce products. You then go market the features and you go market how that company has new solutions to specific problems, right? The whole purpose of features are the features are supposed to solve problems that their ideal customer is facing. And what we're finding now is, And we've been seeing this for more than a year now because AI has drastically increased the output for B2B software companies. They're cranking out so many more features and doing more things with their product than they ever did before that now companies are like, we need to hire more and we need to do more or it's going to be too hard to keep up on the marketing side. The main reason being is they wouldn't have the feature set that their competition would have Which is going to make it harder to market a product. Because no matter how good you are at marketing a product, if someone's looking to get A, B, and C done, and your competitors can do all 3, A, B, and C, and you can only do A, well, it doesn't matter how good you are at marketing. If you don't do B and C, you're going to start losing customers.

Yep. Right. So actually, by the way, go ahead, finish it up.

No, but we pretty much ended there, but it goes to people have to hire. They're now in a rat race and you have no choice, but to just move quicker, which is causing people to have to increase the amount of employees they have for different divisions, including marketing or engineering.

Yeah, I think, by the way, it's because you're forced to launch new features. Guess what? The role of product marketers is becoming more and more important because now it's like every, not even every month or so, you might have every week or every 2 weeks or so you're launching something that you look at how fast Anthropic or OpenAI are shipping new products all the time, or even Google. Shipping new product features all the time. It's happening at lightning speed. And so Neil's point here, by the way, maybe it's becoming kind of the status quo now to have to move that fast and that might burn people out, but I don't know, man. If it's the status quo, someone else is more than happy to take your spot if you don't want to move that quickly. And maybe that's just the reality right now. I don't know. We'll see.

No, I think it is the status quo. We're also seeing engineers being offered a lot more money than they used to, at least the good ones. We're seeing companies trying to poach really hard. I bet you you're going to experience this if you already haven't. We've experienced this as well. And at least for us internally, when I look at our projects like Answer the Public and Ubersuggest, we used to try to release— our goal was at least one major feature a month or one major change. It's gotten to the point where the team's releasing so much on a weekly basis, it's hard for me to keep track of because I'm traveling or doing other things. So it's like, dude, If they're messaging me on WhatsApp on 4 new things that they did this week, it's starting to become hard for me to keep up. And I think other marketers are going to have the similar issues unless you have someone focus on, hey, I'm going to focus on product marketing and just create material around these new features and videos and social posts and do SEO campaigns around it and paid advertising campaigns. But it's becoming overwhelming because of how fast People are moving within organizations now.

Yep. I'm going to show you something real quick. Because we're talking about how fast AI is moving right now, I'm going to show you how we've been communicating faster because of the speed of AI. Basically what we're having people do now is create artifacts. What I mean by an artifact, if you see this on the screen right now, is we're having people create HTML documents. Imagine Neil and I were recording this podcast right now. I can take this transcript or maybe I'm doing Granola, I can take the transcript from that, and then I can use our design.md to design in whatever format I want and make an HTML file that then I can share with my entire team. Maybe I can share with one of the teams, but the whole idea is that we can communicate a lot faster with 1 or 2, 3 slides or so. In this case, someone's showing me the type of ads that they want to run for ChatGPT ads. But this is how we're talking now because it's like instead of having to read the entire transcript, It's hard to parse that information. So if you can have the AI parse it for you, make it into a nice little HTML file, it makes it way easier for you to get information across. And if everyone is aligned with where you're trying to go, let's say you're communicating to 10 people or 100 people or so, if you can align those people that quickly, it causes less problems in your organization. So I'd highly recommend that whenever you're talking to people now, one, have your recorders on and then make these artifacts. And make as many artifacts as you can because it just makes life a lot easier when it comes to communication, at least for us.

Yeah, no, I'm with you. Communication's also really huge in marketing and business. I think not enough organizations spend enough time fine-tuning their communication processes and working together as a team and figuring out how different people communicate with each other and how they comprehend what you're communicating.

Communicating.

I remember Eric used to make me take surveys or quizzes. I don't know what they were called. Personality tests.

There you go.

And he would make friends take them and other entrepreneurs that he knows.

And it would make other people take them too. Yes.

I don't want to mention that, but I still remember that. And with these tests, it's really good. Even though I wasn't a fan of them and Eric would make me take them, he's like, can you just do, just please take it. And he would pester me until I finished it. And then he would send me the link so he would get access to the data. It was great because he would understand how my personality is and the way I think, and he would share how he thinks, and it would allow us to better communicate with each other because we understood what made the other person tick or move or what they didn't like to hear. Like Eric knew that I love data. So before, years and years ago, Eric used to be like, you got to try this, you got to try this, this is a better way. And he noticed that I would barely do anything with any of the information that he passed me. Then after the personality test, he quickly learned that I'm very data-driven and data-oriented. And I'm— I can— someone can change my mind really easily if they have data and facts to back it up. Like, I'm not stuck in my way as long as you have something to prove me wrong. And I'm happy to be proven wrong, especially if what you show me can make me better or make my business grow. So then when Eric would come to me, he would start using facts, stats, and data, even if it was a small sample set, because he knew that it would be more likely to get me to take action.

You know what's funny? The best data is like, hey, do you see this actually makes money? And you're like, oh, it's a little different when you have the money there. But on the note, Neil got smart with that. You actually removed it, so I can't pull it up anymore. So Neil got smart and deleted it.

No, I didn't remove it.

But you're not on my thing anymore, so it must have done something. But what I'll say is the personality test thing is huge too because you know why? Once you start adding it into whatever AI harness that you're using, whatever agent harness you're using, you can then say, oh, hey, you have the entire repository. Hey, today I'm going to be giving feedback to Neil. How does he prefer to receive his feedback? Because Neil and I, one reason why we get along so well is because I prefer to take my feedback direct. No sugarcoating, but the vast majority of people don't like that. They take a lot of personal offense to it, right? And so, but the thing is, when I look at Neil's personality test on the tough meter, he's like 99%, right? So he's very direct, but also his composure, 99%, meaning that, and I've actually never seen, I very rarely see Neil lose his poo-poo, right? Only when he's getting screwed on something will he lose it, but it's very rare. So all that to say is, I know where I can push. I know where I need to let go sometimes. And that's why I think it's worth it to have everyone in your company do it. And now that you have AI, it's good to have all of that stored so you can work better with your team. Yeah.

Although what I've been doing, and this actually could be why my personality test isn't shared with you anymore. I don't know when it was, but let's call it a year ago or 2 years ago. There was some time where we started sending out a mass email to everyone that my information was subscribed to because you get these emails. To erase and delete the data.

Oh, okay. That's what happened.

That's what happened. So I don't know if they actually did it or not. I knew I was getting a lot of follow-up emails from that, but I just kind of ignored the follow-ups. It was just more so—

That was 2 years ago. So I think you disappeared 2 years ago.

That's what it was. Yeah. We're just like, delete, because we don't want people having it. Because what we saw is when your data's out there, a lot of these organizations are getting hacked. Might as well just remove it from any place that it doesn't need to be.

Yeah. Although, I mean, that data's like not that valuable. Yeah, it's like a valuable— Yeah. So, okay, let me pull up this Paul Graham post. So Paul Graham is the founder of Y Combinator. So Neil, I wanna get your thoughts on this. So he says, the only thing worse than having the CEO knee deep in building stuff with AI is not having the CEO knee deep in building stuff with AI.

I look at that as it depends on the type of company you are. Paul Graham typically deals with startups and product people and engineers. So in that case, I totally agree. But for all companies, I would not say it's a good fit to have someone knee-deep just building things in AI. Example of this is, and this is extreme example, let's say you're a cold storage trucking company. I know someone who actually owns one of those things where fruits, vegetables, when it gets transported or food gets transported, a lot of it has to be in different temperature-controlled storage units, and then it has to be delivered to facilities that are temperature-controlled. And then from there it gets distributed to, let's say, grocery stores. Having the CEO of that company being knee-deep in AI, I'm not saying AI can't help 'em. I don't know if them focusing the majority of their time on AI is right for that industry or that job, but there are cases where it is, and I would say there's also cases where it isn't.

Dude, I gotta show you this graph shortly. But my take on the Paul Graham post is this. So one, I think he's saying that, okay, it's really bad to have the CEO knee deep in building stuff with AI because it causes a lot of distraction. It causes a lot of thrash in the organization too, because now that someone like me, especially I have all these ideas all the time, I'm maybe starting to pry into other people's GitHub repos. Hey, how's this going? Hey, I'm gonna ship this PR today. So it does start to cause a lot of chaos. But I think he's also saying for tech startups, but if also the CEO's not super AI-pilled, he's not going to know how to really steer the ship, right? At least in a tech startup environment. Now look, if you're a trucking company or we're talking about an agricultural company, maybe not so much. But my take on it is I think that if you're a CEO and you're driving the ship on the strategy here, you should at least be spending some time, but I don't think it's like, it also depends on the business to Neil's point.

So dude, what's funny on agriculture is actually not to say the CEO, but they're really heavy in AI right now.

I believe it.

Yeah. Because agriculture is just so— and because we work with some of the biggest agriculture companies out there like Cargill, and we work with, I think the number 2 and number 3 player as well. Cargill, most people won't know what type of company it is, but it's an agriculture company. They're privately owned. They do around $160 billion a year in revenue. At least that's what Google says. I think it's one of the largest privately owned companies in the world. There's not even that many publicly traded companies in the world that are generating $160 billion in annual revenue.

Yeah, I think you kind of need— so going back to this, as I was saying the agriculture thing, I was like, I'd imagine they'd be pretty big with AI. But my thing is this, if you're a CEO, you should really understand your space well, but you should also understand how to use the augmentation power of AI. And that's why Paul Graham is saying this. So I do think if you're listening to this, my stance is yes, you should be knee deep in it, but you shouldn't be disrupting your company. That's what I would say. So by the way, speaking of, because we're talking about bots right now, did you see the Cloudflare traffic over time? How bots have basically surpassed humans? Here, check this out.

Yeah, 51% of now website traffic is bot traffic.

Exactly. 57%. Okay. So if I look at the— let me just put this to the last 7 days or so. So the CEO, I think his name is Matt, he said that he didn't see bot traffic surpassing human traffic until maybe like 2027, but it has already happened 6 months into 2026. And so, okay, the question now is like, okay, what do you do about that? We've been talking about this on this podcast since a few months ago. You have to start optimizing for agents, whether that means having really good documentation, whether that means for your products, like Ubersuggest has an MCP, right? ClickFunnels has an MCP and an API. You want to make it easy for agents to find things and you want to make it easy for agents to use things as well because we are now optimizing not just B2B, B2C, we're doing B2A, business to agent.

Yes. We did a webinar on that. I don't know when, maybe 3, 4 months ago on how bot traffic has oversurpassed human traffic. And I think the title of the webinar was how to optimize for bots or like in the new world. I'm messing and butchering the title. We had something that was much more catchy, but the webinar did really well because people aren't thinking about optimizing for bots. But here's the caveat. As marketers, you've been optimizing for bots for a very long time and for the majority of your career, you just don't know it. You optimize for Googlebot, you optimize for Meta and Instagram and their algorithms. Yes, you did also optimize for humans, but you've been working on optimizing for bots for a long time. There's not much of a difference when it comes to optimizing for agents. They look for some of the same things that humans do. They look for some things that may be different. So you got to think structured data, right? So if you've already been optimizing for robots like the Googlebot, or the ChatGPT bot, I forgot what it was called, it's something like that, you should be fine. Just think about what an agent would want to see and how to make the experience delightful for them. Still think about traditional platforms as well and what their bots would want to see and also think about humans as well and what they want to see. Now you're really optimizing for 3 things, the platform bots, Agents, and the third one is humans.

Cool. By the way, speaking of, because we're talking about SEO right now, I think the most interesting thing is Google has just made SEO a positioning game again. So if we look at what Lily Rae here said, so she says, Google just put out its first official article on optimizing for AI search, AEO/Geo. The TL;DR, too long, didn't read, Is SEO still the foundation for AI search? Wow. That's kind of a joke a little bit, but create non-commodity people-first content. Ignore most GEO/AEO hacks like chunking and llms.txt. So the first 2 here, when I say it's a joke, I'm not talking about Lily Ray here. I'm just saying that these things SEOs have known for a very long time that these SEO and AEO/Geo, they kind of go hand in hand. And when you're talking about creating non-commodity people-first content, that's ever since the very first day I started learning SEO, it's been like that. And then all these other hacks over here, they don't matter that much. At least that's what Google is saying, right? With these llms.txt and chunking. That's the first piece. You have any reactions to this one before I move to the next one?

Yeah. And when Lily Ray's posting this, she's not saying this is how you win in this new AEO or GEO world. She's just summarizing what Google puts out. Right? So Google will tell you what they want, their narrative. It doesn't mean that they're going to tell you everything. It also doesn't mean that they're hiding anything either. I do believe SEO is still the foundation. There are some nuances and some differences, but it still is the foundation. And you have to create amazing content that people haven't seen before that they want to engage with. And I think if you do that, it'll start doing really well in the LLMs as well. People really struggle with number 2. I don't think too many people have an issue with the fundamentals of SEO because it's been around for a long time, but they do struggle on the content side because content marketing has been around for so long. It's more and more competitive and there's already a lot of information on the web. So you have an issue of regurgitation and people are tired of the same stuff over and over again.

Cool. And so by the way, Tim Soulo shared this and this happened June 2nd. He got a million views on this. So in the last 6 months, Ahrefs, we analyzed over 1 billion data points across 14 studies. And here's what we learned about AI search optimization. Number 1, best X blog listicles are the single most prominent content format cited by AI chatbots. They make up 44% of all pages cited by ChatGPT specifically. Number 2, 67% of ChatGPT's top 1,000 citations come from sources marketers can influence. Number 1 is Wikipedia, 30%. Homepages, 24%. App stores, 6.6%. Only 32% are influencer content like educational pages, reviews, news, and blog posts. I'm going to read one more here and then we can talk about this.

Read them all. Read them all. There's like 10 here.

Okay, I'm going to read 3. You can read the rest. 3, 28.3% of ChatGPT's most cited pages have zero Google organic visibility. These pages get cited repeatedly by ChatGPT despite not ranking in Google at all, a completely separate discovery layer. Okay, that's interesting. Number 4, ChatGPT only cites about 50% of the URLs it retrieves. It fetches dozens of pages per query but uses half as background context without attribution. This means that being retrieved and being cited are very different things. Number 5, before I turn it over to Neil to read the next 5, Number 5, adding schema markup had zero impactful, meaningful impact on AI citations. AI Overviews actually dipped 4.6%, while AI Mode 2.4% and ChatGPT 2.2% showed changes indistinguishable from zero. Okay, let's get to the next.

Number 6, YouTube mentions have the highest correlation, 0.737, with AI branded visibility out of all the factors we studied, including all the conventional SEO metrics like backlinks, page count, domain rating, etc. This held true for both Google-owned and OpenAI products. Number 7, AI overviews reduce clicks to the number 1 result by 58%. That's up from 34.5% just 10 months earlier. This trend is accelerating. Number 8, 99% of AI overviews appear on informational intent queries. Transactional, navigational, and local searches are almost entirely AI overview-free. Shopping triggers AI overviews just 3.2% of the time. Number 9, for a given search query, Google's AI mode and AI overviews reach the same conclusions 86% of the time, but cite almost entirely different sources. Only 13.7% citation overlap, or 13.7% citation overlap. Number 10, AI overviews change every 2.15 days on average, with 70% of the content differing between consecutive observations, but semantic similarity stays at 0.95. The words, the sources, the entities constantly shuffle, but the actual meaning barely moves.

By the way, for number 8, it's 99.9%. So basically 100% of AI overviews appear On informational intent queries. I think this is good data, by the way. This is really good. So where do we want to start first? I think this was a surprise to me. Number 5, adding schema markup had zero meaningful impact on AI citation. So all the stuff like AEOs, SEOs like to talk about, that was a little surprising to me.

Well, here's clarification here, Eric. He's talking specifically about AI citations. He's not saying it doesn't impact them pulling your content. This is just citing.

So that's why it was surprising me.

I didn't read it right. Okay. Right. Because, but we're all optimizing for citations. We're not optimizing for them pulling from you and not citing you. Yep.

Number 6, YouTube mentions had the highest correlation with AI brand visibility out of all the factors we included. So I mean, it depends. I mean, with Google-owned products, and it's also true for OpenAI products. And I think we revealed this in the last HubSpot that we talked about. I think we pointed out the graph where it's like, The top 3 sources are, they're interchangeable, like the rankings shift, but Reddit, YouTube, and Wikipedia. So that makes a lot of sense. What else is interesting to you here?

So number 8 is interesting for me because I think it's going to change. So this, Eric and I know, in which AORVs really only appear on informational intent queries. It's the number 1 area where they appear. You don't really see them much on transactional navigation on local searches, which we've seen. But here's the kicker. Tim mentioned shopping triggers AI overviews just 3.2% of the time. I'm not saying AI overviews are going to appear more for shopping-related queries. I do believe shopping is going to change because of how Google released their new search features with agents. Eric, if you want to open up a new tab and we can show people on the screen here on what we're talking about.

Which tab? What do you want me to open up?

Just open up a new tab. You know, from the Google I/O Marketing Live in which they talked about the new search tab and shopping features?

Yeah.

You see that article? And they use the dresses example. Find me this dress, but in this color. And they uploaded 2 images and then Agent just goes out and finds it.

Find me a blue dress for a wedding. You mean like dragging something in? But it's okay.

So go back really quickly.

Uh-huh.

Okay, so there's a plus button that's being highlighted, right? You saw the Google Colors right there. You can actually upload images. So you can upload 2 images. You would upload the first image. Eric's using a blue dress as an example. So you would upload the dress you're looking for, like the style. The second image that you could upload is maybe the color of the fabric. And you can say, find me a dress that looks like this but in the color red. Or you can upload a picture of the specific fabric you want. And Google will just go and find it. And you're seeing it right here on the screen. So Eric uploaded an image of a blue dress and it's showing you all the places with a similar style that sell that dress.

Oh, that's pretty cool. You want me to buy you a dress?

I think I'm good on the dress end, but I appreciate the offer. Anyway, this is, this is how I think shopping is going to be affected, not necessarily by AI overviews, uh, but the 10 blue links. Are— and even the shopping category is going to be really affected by people just typing in exactly what they want. Not only can you describe it now in words, you can upload multiple images, not just one. You can say, I like this dress style in this color, but I wanted this dress length. So 3 different images: one image with dress length, one image on the color, one image on the style. And it will go scour the web and find you all the ones for sale that might be right for you. And you can even tell it the size you want and the price range, and it'll go figure it all out for you. That's going to change shopping.

It happened very quickly too, so that, that, that's awesome. Um, cool.

A few seconds. It didn't take long at all. It was quick.

Yeah, I think so. By the way, I mean, I just want to call out the fact that this post— because I, I think Tim does well on X, um, but I think this is one of his, um, His better ones. When you're super data dense like this and you have a couple of interesting points here, this is very shareable. You got 7,000 bookmarks on it. I just retweeted it myself, a million views on it. I think that's great.

Yes.

Cool. All right, let's move to the next topic here. So Neil, let's just talk about your— every time you travel or I travel, we should talk about what we're doing. So what are you doing in Jakarta right now? What are the lessons from this week? What are the conversations that are happening right now?

Tell us, fill us in. Sure. Uh, so a few things. Um, I left on Sunday— no, Monday. Yeah, I think I left on Monday. Left on Monday for— from London, or LA to London. Landed in London Tuesday night. Uh, spoke at South by Southwest Wednesday. Uh, Wednesday night I flew from London to Jakarta. I had to go through Dubai, so it was a little bit of a long flight because there's not Direct flight. Emirates is nice, and it was really cheap, dude. How much do you think my round trip from London going through Dubai to Jakarta first class and going first class from Jakarta to Dubai to Rio was in cost? The Dubai to Rio Jakarta to Dubai, I think, is around eight hours. Dubai to Rio is fifteen. So I got both those legs. Dubai, Rio, uh, I mean, yeah, Jakarta to Rio and London to Jakarta, right?

I'm gonna say the way you're saying it right now, because it's a, it's Emirates first class as well, or you had pretty good first class, I want to say $7,000.

Close, $8,000.

Yeah, that's a good deal.

I don't— yeah, you're talking about normally a plane ticket that would have been $25,000, $30,000.

Exactly. Yeah. And so that's, that's a good deal.

It's a good deal. And sadly, I think it's because of war and not everyone wants to fly through some of these regions.

Nobody wants to go through Dubai right now. You crazy?

Yeah. And I think that's why it was so cheap for a first-class ticket. Um, so why am I doing all this traveling? London was South by Southwest. We tend to do it because of more branding for our European agencies. Uh, Indonesia, we have an APAC division, but we don't do tons of business in Indonesia. Uh, the GDP per capita isn't too high. I'm not trying to talk trash, I'm just telling you the real reality. So we just don't see companies spending as much in marketing. And I haven't actually flown to Indonesia before, but my team's talked about buying an agency in Indonesia multiple times because they're like, oh, it's going to be the next big region in APAC. And agencies have been saying that for a long time. And I was like, you know what, I have the opportunity to go Let me just go. There's going to be 5,000 people at this conference, so I'll end up seeing what they're doing, seeing some of the companies. I'll meet a lot of the sponsors. I'll get to know the lay of the land. And then from there, I'll head to Rio for Web Summit. Rio for Web Summit is one of our better events of the year that we tend to speak at, because when you look at most Web Summit events, they bring in massive crowds and they come from all parts of the world. Um, the Portugal one's really good. The Canada one, it used to be called Collision. They changed it to Web Summit Vancouver. I used to prefer Collision Toronto. It was still owned by them. Uh, it was bigger than Web Summit Vancouver, although they're probably working on building it up.

Why aren't you going to Cannes?

Cannes is more creative. We do creative stuff, like we did stuff with Formula 1, WNBA, Trojan condoms recently. The list goes on and on on the creative campaigns that we do. But just being very transparent in the agency world, for anyone who has an agency or is looking to hire one, there's not much money to be made on creative. Yeah. It's not because of AI either. Even after AI and before AI, there still wasn't a lot of money to be made on creative. And the reason I say that is creative is in the eye of the beholder. So if Eric hires NP Digital, to do his creative, I can say we created something amazing. Eric has his own taste. He could say he dislikes it or likes it and wants another version or tweaks or whatever it may be. And because of that, because it's in the eye of the beholder, and when you work with large organizations that have tons of people who are meddling and getting involved in the creatives, the margins just suck. And it's not just us. Most creative agencies don't have great margins. But you have to offer creative solutions to get the rest of the contract, which is why we acquired a creative agency years ago. But the reason we're not at Cannes is it's a low margin side of the business for us. And most of the people who are at Cannes are going there to party and have fun. So you tend not to pick up a ton of business from what we've seen.

Exactly. So I was at an event last week and then a couple of entrepreneurs asked me, oh dude, Cannes makes a lot of sense. I'm like, No, it doesn't. I'm like, I'm not there— like, more power to you if you want to party. I'm just not there to party. Like, if I'm going to travel, I want to— either it's for vacation, I've been there to like, you know, have good food, or I'm there to do business and then meet with some people and then get out of there, which is what Neil does 99% of the time.

So yeah, the flight attendants are, uh, yeah, flight attendants think I'm crazy because a lot of times, A, I think I fly more than most flight attendants I'm pretty sure on that, the amount of hours that I clock. And B, I see a lot of the flight attendants like multiple times. Like even on my trip to Jakarta from London, you know, flight attendants said, oh, goodbye. And I was like, oh, I'll probably see you tomorrow. And they're like, you're leaving tomorrow? I was like, yeah. Same with when I was in Hong Kong the other day. I left on a Monday afternoon. I think it was a United flight, like around 10, 11:00 AM. Got there Tuesday sometime their time. Wednesday I spoke, did my meetings, and I flew right back. And I was home Wednesday night because of the time change, and the flight attendant's like, you're on our flight again? Um, but that's normal. And then I was home for a few days, and then I was off on the road again.

I'm surprised you don't get sick more because that— this stuff really screws with your immune system.

Dude, you want to know the secret? It's changed everything in my immune system. Vitamin D drops. I'm not trying to give health advice to anyone. But, uh, Jacob, you know Jacob who has the—

yeah, yeah, I talked, I talked to him earlier this week. Uh, I'm getting some hooked up hookups on peptides. Shout out to Jacob and Abud. Abud just did the Andrew Huberman podcast. I'm getting some, some peptides.

Yeah, so Jacob ran some tests on me. Uh, we've been paying him for many, many years, and he's just like, dude, you get sick a lot. And I was like, are you asking me a question? He's like, no, you get sick a lot. And I'm like, how do you know? He's like the vitamin, I think D level was really low in me, and he's just like that's the number one reason we know if people get sick a lot, especially when they're traveling. He's like you need these vitamin D drops. He doesn't sell them. He's like just go on Amazon and buy them. Here's some few brands that you can pick from. I bought one, and I've been using it for years, and I take a lot of the drops, maybe more than I need to. Again, not medical advice here, and that stopped me from getting sick. I would probably say. 9 out of 10 times.

You got to be careful because I, I did the same. Like, my vitamin D was really low, but you can also go too high on it too because— so you gotta, you gotta, you gotta do your, your blood draws. But, um, I will say to your point, Neil, um, I was low on vitamin D maybe 3 or 4 years ago. I haven't really been sick for 3, 4 years. Um, so, you know, there's, there's a trend there. Not health advice, guys, but, um, it's just what worked for us. So I want to switch gears here and talk about really AI is cool and all that, but there's actually a bottleneck. Most companies are failing with AI if they don't really touch this. You can see, Neil, my good friend here, Gary Tan, said this is the actual bottleneck. The models are actually smart enough already. What is missing is a company-specific context locked in senior people's heads. Whoever cracks knowledge extraction at the company level unlocks the rest. As you work on this, please consider using GBrain. That's his That's his product, right? But what really he's calling out here, what Gary Tan's calling out here is that Neil, for example, myself, we started with our organizations. You started since the beginning. I came in a little later for my organization and I took it over. But point is you and I, we have the most context. But not only that, your CEO, for example, or other executives at your company, they know the most and they can see more of the battlefield than most people. So how do you translate all that context into one area that maybe some shared context that other people can access? Because if they can pick that up, you're able to move a lot faster, right? So I think if you can extract knowledge at the company level and safely where there's appropriate permissioning and all that, I think you're going to be able to move a lot faster. And that's why, Neil, you see so many people right now building this whole company memory or company brain stuff. I think everyone's trying to crack it right now. And there's a lot of horizontal solutions. But anyway, I think it's something that most people should look at if they wanna be able to move faster.

Yes. So that was his company. Did he end up raising money for his company through Y Combinator?

No, he didn't, 'cause it's open source.

Oh, cool. So he's just giving it away.

He's giving it away, correct.

He's probably already rich from all the Y Combinator money. So is Paul Graham. People are like, ah, does Paul Graham have a lot of money? I'm like, it's not disclosed, but I bet you he has over a billion bucks.

Yeah. What is it? Airbnb, Dropbox, like what else is there?

There's so many, dude. And there's so many companies that we haven't heard of that are in B2B that are just worth a killing and were in Y Combinator.

Yeah. Anyway, it was a good model for them. But anyway, just make sure that you pay attention to that. If you don't, I think most companies will fall behind on that side. So, okay. I think Neil, you'll like this piece because we've kind of talked about this a little bit. I'm curious because I'm in these AI chats as well. Now you see executives talking about, How people are starting to rationalize their token cost, right? So if you look at this over here, so this is the founder of Lindy. So pull the trigger today and switch to 100% of Lindy traffic to DeepSeek V4 churning from Anthropic models, right? So the Claude models. This saves us millions of dollars and we're actually seeing an increase in performance on many core use cases. This has been transformative for the business. And then so Chamath here, He tweets this. He says, if chapter 1 of AI was all about broad open-ended experimentation, chapter 2 may well be about realism and rationalizing the cost of chapter 1 and what is sustainable moving forward. And I think you're seeing that with your clients. I'm seeing that with my clients as well, where they're just spending an arm and a leg. You see Microsoft like, oh, we got to cut. Uber's like, oh, we got to cut down. It's not to say this stuff isn't productive, but you can't just spend and spend and spend. And so Sam Altman even said earlier this week, this is the first time they're seeing people tighten their belts.

Yes. And I think it's going to continue because these things are just so expensive. Uh, Anthropic signed a deal with SpaceX. I think it's like $15 billion a year, I believe it is, for that AI.

I think they're paying them $1.52 billion a month, something like that. So yeah.

Okay. So, uh, call it around $19-ish billion a year or whatever. That's going to cost SpaceX stock to be even worth more once they IPO.

Absolutely.

But we got an email from Claude a little bit ago, or Anthropic, which is the parent company name. And our account manager there for our enterprise suite was helping us figure out how we can end up reducing costs and integrate them more. So I think they're already on top of it because they're seeing pressure from organizations. And we've been talking about this for a bit. This stuff is great, but there's a cost-benefit analysis. If you can get 5x more done, but 5x more done increases revenue by 3%, but your costs go up another 15%, is it worth it or not? And you have to make the calculation.

Yeah. So what I would say is this. I think the fact that they moved to DeepSeek, I think DeepSeek, you can pay for it or you can host it locally. I have Quinn on my computers over here. So there is a trade-off to all this. Obviously my team can tell immediately when they get degraded to a lower level and then they start complaining. But I would also say that Cognition today, you should take a look at this, Neil. This just occurred to me right now. Cognition came out with a new way of measuring your token effectiveness. Okay? So check this out. There's a new AI productivity guarantee, and I think we're going to see a lot more of this. I'm pulling this up right now. Cognition says AI should earn its keep. Introducing the AI productivity guarantee. If Devin, which is their product, delivers less engineering value than you're paying for, Cognition will fund your usage until it does up to $10 million.

Wow.

That's a good guarantee, man. That's a good offer. It's time the AI industry for the AI industry to stop maximizing tokens and start maximizing productive output, which is what we've been saying for a while. It's not just about token maxing, it's about output maxing. We build a system to estimate whether an agent's output was actually productive, and if so, how long a human engineer would've taken to do the same work, validated on real engineers' time estimates working in enterprise code bases. They do a blog post on this, but I think you got to go in this direction if you want to stay alive. Because these other open-source models are catching up too quickly.

Yeah, and I think these guys, I don't know which models, but companies are not only becoming more careful, we're hearing now from companies globally because we've all heard the US versus China thing, and a lot of US companies are saying, oh yeah, we want US-backed or US-built or whatnot. We got to be careful with with the Chinese products. Uh, I'm just repeating what I'm hearing. We're now seeing a lot of companies when it comes to some of these AI models, whether it's for their engineers or whether it's for the marketers, a lot of them be like, oh, these Chinese models, these ones are a fraction of the cost, let's just try them out. Who cares if it was created by China or someone else? They're like, let's just save the money. And I believe whoever is going to have a good enough model that is cheap enough to help with marketers In the end, it's a lot of it's going to come down to price. It's like my team is on Microsoft. Why are we not on Google? I remember when we made the switch to Microsoft, so many people in the organization, this is years ago, was pinging me. They're like, Neil, we don't want to do this. Come on, can you convince some of the others? I stayed on G Suite because my email is neil@neilpatel. My team has an NP Digital email, so they all had to switch. I didn't switch, so I didn't have the problem. I let the company do what was best for the company financially. And you know what? The team got used to Outlook and they use it. And that happens for a lot of stuff. You know, when I go speak at some of these Fortune 500 companies, I ask them, what LLM models do you use? A lot of them are Microsoft's models. And I say, why? They're like, it's included. Same with people who are on Google's G Suite, like Gemini. It's included. It's easier. It's cheaper. Companies are here to turn a profit, especially publicly traded ones. Why wouldn't they want to save money?

Yeah. I mean, my take on all this stuff is I think it's cool to play with the open source models. It's cool to play with the frontier models. You have to figure out what works for you. I think this output maxing thing is just going to become a lot bigger. If you can, by the way, if you're a services company and you're figuring out how Because we talked to a company this week and the only thing they're focused on, Neil, was actually output maxing. They're like, dude, we're spending so much on tokens now. How can you help us optimize that? So we've said this, I think a week or two ago, that's going to become a service. And by the way, a lot of things that we never would've thought would become services are going to start to become services. So that's why services is not going anywhere. I think agencies, as long as you're on top of this stuff, you're going to be good for a very long time. So maybe the last episode we can do here, Neil, Is on this piece over here. Oh, this one's cool. So do you want to hear about the podcast mention that drove $29 million in revenue?

Yes.

Okay. So this is from Chamath. He got an email. So, hi, J Cal, Chamath from the All-In Podcast. 55 days ago, you had a 4-minute conversation about EAs and Athena. So Athena helps source executive assistants. Since then, we have closed $29 million in new revenue, seen a negative 53% decrease. So a 53% decrease in blended CAC, cost of acquiring a customer, and placed 683 Athena executive partners, 683 new jobs. And so I'm sure he probably has a stake in this company, maybe not. But all that to say is whether you think about a mention on Joe Rogan, for example, or the All-In podcast where you have a lot of executives listening to it, Or you even think about what's the reason? Oh, Jacob's partner, Ebud, on Andrew Huberman's podcast. It's a matter of not just finding the reach, but finding the relevant audience, and then you should be able to crush it. Because now I think that peptide company, by the way, they're going to crush it.

Yeah. I saw this Athena one and I was like, all right. It didn't say it explicitly, but I'm assuming They were mentioned on the All-In podcast and from there it drove the revenue.

Correct. So I had listened to that episode and that's what prompted me to sign up again and see how it goes. But yeah, and speaking of these executive assistants, so David Sacks, another member of the All-In podcast, I'm going full All-In right now, Neil. So David Sacks, Apollo's chief economist, zero evidence of AI-related job loss, what we talked about earlier. So I just want to pull up this data and then we can move to the final topic here. So, zero evidence of AI job-related losses. The chart below shows the weekly ADP employment data, and there's zero evidence of job losses because of AI. Instead, many firms are hiring AI implementation experts, and the data center buildout is building, putting upward pressure on salaries for AI experts and on prices of semiconductors, equipment, and energy. The bottom line is that AI spending boom is stoking both employment and inflation. Wow. Okay. As a result, nonfarm payrolls for May could come in significantly higher than 95,000 expected by a consensus. It is Jevons paradox playing out in real time. Cheaper technology is creating more demand and more jobs. Look at this over here. This is the weekly employment starting from, I guess, last July leading up to now. Pretty crazy, huh?

The one thing I would say about the AI jobs, you are seeing a lot of narratives about people losing their job because of AI and companies like, um, Atlassian, Block, which is Jack Dorsey's company, saying that they're laying off people because of AI. I don't believe it. I believe a lot of these guys are laying people off because they're not getting into the growth numbers they want or their stock price went down, not because we're more AI efficient. And at the same time, from a marketing angle, imagine going to all your shareholders and using the message, marketing message of we're laying off employees because we're not growing as fast as we thought. And we're not hitting numbers, so we're going to go and cut costs. It's just going to hurt your numbers even more. It's easier to use the AI narrative and spin your marketing message than it is to just say the truth.

I would just also say one more thing around this. I think this proves that we talked about short-term job displacement. I think it's going to happen in some areas, but I think you're going to see more task losses versus job losses. A lot of tasks that we had to do before, we're no longer going to have to do because we can just have the agents handling it. And I think, again, it's task losses before job losses. I think that's going to be how it goes for the next couple years. And final thing, Neil, because we've been talking about so much about AI, I feel like we should talk about some boring channels that still win in 2026. So these are some boring channels in marketing that still win in 2026. I'll start with number 1 over here, direct mail, meaning mail that you, physical mail that you get. If I mail you, Neil, if I give you a UPS, if I FedEx something to you, you're damn right you're going to open that thing, right? It doesn't matter what it is. You're going to open it because it seems like it's important to you. So direct mail, what's old is now new again.

Yes. And number 2, email and SMS. I'll combine them because some people just do SMS. Some people just do email. Some people do both. It works. Emails are still getting opened. They're still getting clicked. It's a very effective channel. I don't know why people wouldn't be leveraging email and SMS. SMS, they'll be like, oh, AI's going to just summarize the emails. We're still seeing it work. I saw an interesting stat published by HighLevel. Out of the emails they send, their emails get a click rate of roughly 6%, extremely high. People wouldn't be clicking 6% of the time if emails didn't work.

Yep. Last one we'll go with here before we take off here and let Neil go to bed. Live events, dinners are not going anywhere. Conferences are not going anywhere. In fact, they're gonna become more valuable. Relationships are not going anywhere as long as humans are around. And I think they're gonna, it's gonna be harder and harder to come up with relationships. 'Cause a lot of the young people we talk to right now, the reason why they're going to college is 'cause they know it's gonna be harder and harder to form relationships and they wanna form those relationships. So whether you can create dinners, happy hours, things like that, you can go to events, you can be that super connector, that's gonna go a long way. So that being said, go ahead, Neil.

Let's add number 4, last one. You guys all know this channel, it's SEO. People still evaluate it by clicks. I'm telling you, if you look at your data and you do brand recall studies and you survey people on how they found you, even if you're getting less clicks, if you can rank in the organic listings and you can start ranking, being mentioned in the AI overviews, even if you're not getting the clicks People are seeing your brand, and if you go and look and ask them how they heard about you, you'll start seeing a lot of revenues coming from people seeing you on Google but not actually clicking on over to your website because you're being mentioned in the AI overviews.

Do you, by the way, last 12 months, do you think you're getting net newly— net more leads or net less leads or the same?

Net less leads, net higher quality leads substantially.

Yes. Yes, that's so weird, but yes. Anyway.

But it's not weird because what would happen before is someone would do a search, they see 10 blue links, they may click on 4 or 5 of them, fill out their information. Now people are going to traditional search, and you have to remember, in traditional search there's AI overviews, there's AI mode, or they're going to ChatGPT. They're asking all their questions. The LLMs are identifying which agencies that they should contact So they're going and reaching out to less, but the leads are much more qualified.

Yes. I'm just saying it's weird for you and me because we come from the past, right? Where it's like, I'm just not used to it is all I have to say. Anyway, thank you all for listening. Hope you enjoyed it. Please don't forget to rate, review, and subscribe.