Eric Siu and Neil Patel argue that AI's real competitive advantage isn't smarter models but managing permissions and governance for internal agents. Siu's team spent $7,500 on tokens to save $500,000 to $1 million, a 66× ROI, using AI for ad creation and lead generation.
Chapters
Anthropic sees a moat, Neil. So what do you think this moat is if you were to guess on this moat? I don't— this is 2 up 2, so I don't think you're going to get it. So try it.
No, I don't think I'm going to get it. I wouldn't say it's enterprise or adoption or anything because they're already— all the enterprises are adopting whatever platforms they want. I don't know.
So this is interesting. This guy, Jaya Gupta from Foundation Capital, so he tweeted this. So Anthropic sees this moat. The scarce asset in enterprise AI is shifting from intelligence to permission. And we're seeing this right now, Neil. So same thing, you're getting demand for agents. We're getting a lot of demand for agents and we're building out these revenue agents right now. And what's happening, Neil, is our agents, because my team wants to have agents like mine where they can ask it to do whatever and it has a lot of content, it's connected to all the data that we have internally. And there's like all our SOPs, all these things, right? And what I'm realizing is that permissioning is a very big deal. So obviously you shouldn't let everyone in your company see all of your QuickBooks information or all compensation information, right? You shouldn't have that. And so, models are now good enough. The harder question is whether companies will let them act inside real systems. So, will you let them write code? Will you let every one of your team members merge changes with your code, right? Or touch infrastructure or trigger actions? That is a big deal. And you need to have this governance in place. And then some companies are launching this right now, like Zapier just launched something in terms of governance. And so I think you're gonna have to piecemeal a lot of these things together. We're doing the same thing as well, but that's why we're— we slowly started to test this on ourselves first with our single brain unified intelligence, and then we're slowly rolling it out to clients too. But they're all— Neil, I'll say all of the clients right now, they're all on pilots, um, because we want to make sure that we're building with them every single week because every client has different needs. Um, and that's why I'm like, I'll tell you, like, having to get SOC— getting ready for SOC 2 compliance, like, all these things are a huge pain in the butt, as you know. But we want to make sure that we're building these things right, as I'm sure you guys are too. So governance matters, permission matters quite a bit. Otherwise, it's going to become a poo-poo show at your company and you don't want that to happen.
Yes. And also, when you are leveraging some of these marketing tactics, let's say people are going to— marketing agencies are creating agents for others, or you're going to ride on this wave. Keep in mind that if something has been out there for 6 months or a year and is already popular and like, man, I missed the boat. If you switch the marketing angle to focus more on enterprise, you'll be shocked at how much slower companies adapt and you can penetrate. And then if you're like, well, in the US they're already starting to adapt, that could just be in the US. And a lot of other countries are much slower to adapt than the US because a lot of this technology gets rolled out from San Francisco to the whole US and then to other countries. So what you can do is say, hey, I'm going to end up creating this for SMBs or enterprises in Portuguese or in Hindi or in whatever language, and then adapt for those countries that you know will eventually follow through, but they're not there yet. Yeah.
Remember we talked about this, Neil? Institutional AI versus individual AI. So individual AI could be like you using ChatGPT to help it figure out recipes for you or plan your travel or whatever, okay? Institutional AI is, for example, Neil and I are both marketers. We both know quite a bit on marketing. I won't say we're the best in the world, but We have a lot of SOPs, we have staff that help us. And so what ends up becoming more important is the skill.md files that you have throughout your whole organization, because that's how these agents read this stuff, right? It's all these skill.md files. So the more that you have and the deeper they are, the more of an advantage you're going to have. And you want to go institutional in terms of deep in one area and not just try to be everything to everyone.
Yes. Did I give you the example of how AI really messed up a family vacation for me?
Yeah, you told me just like the 3 times where it spent like $7,500 over or something like that.
No, no, how it put me on a cruise ship. Did I tell you that?
No, no, no, go.
How Michelle wanted me to spend more time with the kids. She accepted, as my wife, she already knows I work a lot, but it's more so the kids are starting to cry. Like even my daughter just said, you haven't come to any of my school functions. I'm like, I volunteer during lunch and stuff. She's like, but you haven't seen my open— you won't come for open house and see the homework I did.
You haven't seen my open class.
And then, and I'm like, honey, I'm so sorry. I had a conference booked 3 months ago or some meetings in a different country. I can't just fly back from Italy and then fly back over. And so my wife planned a Disney cruise, okay, using ChatGPT, shitty internet, coordinated with my assistant so she knows when my calendar was free and all that kind of stuff. So the Disney cruise from Vancouver to Alaska, ChatGPT ends up breaking down like, oh, here's a perfect date in June. Keep in mind, I had a few dates open for many months or a few weeks open. And it tells my wife how June is the best time. All right. So then we tell some friends who have kids similar to our kids' age that are in that area, like, hey, do you want to join us and whatnot? Hands down, everyone we told— some of them may end up joining us, but hands down, everyone we told said, why are you going in June? July is way better. June is not a— good month to go on the cruise. So then my wife goes back to ChatGPT. Hey, you told me June is the best month to go. Everyone's telling me July. And they're like, oh, you make a good point here. But the reason I picked June is when you look outside, you can still see a little bit of snow at the top of the mountain, which looks beautiful. So that's why we say it's the best month. But then you can't do crap outside because of the weather and it's not conducive. And my wife's like, you've picked a crap month and wasted— made us waste money on a non-refundable trip. And not just a little bit of money. Flights, I think, were like $4,000. The cruise, $20-something thousand. I do not know how the heck people can afford this Disney cruise. I was pissed off at the price. Like, people don't have options. This isn't a fancy room. This is like a Motel 6 looking room. I'm like, how the heck does it cost this much money? They just charge that much and get away with it.
But— ChatGPT, this is how ChatGPT screwed you out of $30,000.
Probably more, but yes, to start off with, and it picked the wrong date.
Probably fifty.
And I'll have probably fifty. And then I'll have forget the cost on that. I'll have crappy internet for the whole week, so I'll be very unproductive.
That will cost you even. Yeah, it's gonna cost you way more.
Yeah, but I get that one because I've been working so much, and Eric sent me amazing X quote or quote from X from this. Guy talking about the net worth versus spending time with kids and how you can't take it with you. And I agree with him and I've been trying to make it better, but I actually found I've probably traveled more this year than last year and it's probably worse.
I was just going to say before you said that, I was like, I need to remind Neil to reread the IRR meme. So the whole idea with that one is nobody, your daughters are, and your son's not going to care at age 21 how much your IRR was at 31% versus 25%. So you can spend more time with them, right? I was going to say, Neil, you've actually regressed this year. You need to reread that.
Yeah, it's gone a little crazy. You've done really well. You've cut back traveling a lot, if I'm not mistaken. Like, I've noticed way less travel, and you spent a lot more time on operations on your business, like actually doing the work instead of going out there and trying to build a brand, collect leads, and all that kind of stuff. What has been— out of curiosity, if you had to put some tangible numbers or percentages, what would you say? Rough range, because I know it's too hard to exact, but you probably have a rough idea because I know you would have already adapted if it was hurting you.
Yeah. So here's what I can share. So I will say, I was, Neil, like we can do a quick comparison. At the high point, I was probably traveling half of the year. So 140 to 150 days, sometimes even like 170 or something like that. I think you're probably on track for probably a little more than that. It sounds like.
I think I'm on track for like 65% to 70%. Okay.
So Neil's worse than I am, but I used to be kind of around that. What I found though is that me focusing on my business, I'm working harder than I ever have. And so usually Neil and I would be focused more on, or at least for myself, I'd be focused more on strategy and driving distribution. And this to me is part of driving distribution, but I'm doing more than I ever have done. So I think now, Neil, you can't just be a strategist, you have to be a doer as well. And then you and I have always, we're willing to kind of pick up a shovel or like clean the bathrooms if we have to do it. And so I would just say that I've been more productive than ever because I'm not distracted by travel. And I've been, I'm able to quickly meet with people more. I'm able to quickly just, you know, huddle them or something like that, right? So communication cycles are a lot faster now. And I would just say that without me focusing on this, I wouldn't have gotten the SingleBrain thing out, which I think SingleBrain is the, you know, AI marketing services company of the future. And that to me, like building that future is very important.
Yeah, no, I think that's a great use of your time. But let's go back to your organization single grain, okay? I know for a fact AI has helped you save in addition to whatever you spent. If you look at total savings, it saved you probably over $500,000 to $1 million and told you where to save money without reducing headcount, like without having to fire people or anything like that. It's actually saved you money in a positive way.
Yeah, so what I can report on is, so I actually made a video on this yesterday, Neil. My agent, just on cost savings alone. So I spent $7,500 on tokens last month. Okay. And, but the savings on that, if you do $500 grand divided by the $7,500, that's a 66.66x return. Okay. Just on that alone. And I made those cuts within 3 days. So we cut a ton of software costs. We did cut contractor costs, but we didn't cut— what I can proudly say for right now is like, we haven't— from that, we haven't tried to cut like full-time costs. Okay.
Yeah.
And that's, that's a good thing. I think it's like a lot of people are saying, oh, we're good, we gotta cut like 20%, 50% here. Like, I talk about it a lot, but that's not something that we're actively doing right now. That also on top of that, Neil, the way my team uses it now, like, sure, I spent the $7,500, but that was the toll booth to get everyone else using it like this. So everyone else is getting a lift. They're like, oh, I built this over here. I built this. I saw someone on the team build this. We have something now that can crank out like 400, 600 ads. And we're hooking it into SingleBrain. We have like an API for that. And then end-to-end, having Alfred, our agent, build out cold email infrastructure, which we've talked about a little bit. So, it's building out all these end-to-end things. Had I not spent the time and the money on that, we wouldn't have been able to show the team that we can do it and unlock it for everyone else. And that to me, like I think we can see the gains on that. Come back to me in like 6 months or so, but I think we'll have harder numbers there. Also, Neil, the, the single brain website generated in 2 hours, the enterprise leads coming through that right now, like that's real. That's real pipeline. Quick plug. If you want to win at AI SEO, then you got to check out ClickFlow at clickflow.com. It creates content that actually sounds like you, not AI slop. And there's a free 14-day free trial.
Okay.
Back to it.
Yeah. Yeah, no, I'm with you and I've been on the opposite. I haven't been. in the trenches as much as Eric has this year in business. My role this year, if I had to say, I've been spending, and even a lot of last year, but it's gotten worse this year. I spent a lot of my time networking and doing meals and doing presentations. When I say presentations, I'm not talking about going to a conference, even though I do that quite a bit, and I just finished one up earlier today. I'm talking about going into companies and meeting their executive team, doing presentations internally about business strategy, marketing strategy, How they can adapt, what happens when low-cost competitors enter the market, how should they adapt their marketing messages and pivot, still try to maintain and grow market share. And we'll do lunches and dinners with these people and get to know them. And the reason this is important is I'm playing like a game of whale hunting, right? I don't actually hunt whales, but you guys get the point in which—
You call this spearing, Neil.
Spearing, yeah. What I'm trying to do is land multi-7 or 8-figure deals. And it's a lot of those, it doesn't matter how good you are with AI or marketing or any of that stuff. A lot of that comes down to relationships and wining and dining and meeting and hanging out and just getting to know people because people don't like signing big contracts with someone they don't actually know. I'm not talking about, oh, I see Neil and Eric on the internet type of know. I'm talking about, Sit down and have a drink, even though I don't drink alcohol, but like, you know, have a meal or break bread or whatever you want to call it. Like really getting to know the person, their family and what they do for fun and all that good stuff.
You know what's unexpected, Neil, is the pilots that are coming through right now for Single Brain. A lot of them are, to your point, relationships. And a lot of them are through what? YPO, right? Even though I talked about— but that YPO is networking. What Neil's talking about is a form of networking. I think that's very Everything comes down to relationships at the end of the day. And I think it is high leverage for Neil to be doing this. And Neil, I've got a question for you. What happened to that 9-figure deal?
The 9-figure deal did not close. Although I got a verbal on an 8-figure deal yesterday. So that should close. So that's not bad. And when I say verbal, it wasn't verbal. We were going over the last red lines on a phone call.
It's basically done. Yeah. So what Neil does, you want to remind them what you do with the outreach? Because I think it's something everybody can do. It's free. You might as well do it.
Yeah. So what I really think about is who am I connected with that knows other really influential people at other companies? And I will try to convince the other person because I've built relationships with all these people that I'm connected with to go and talk to the person that they know at a larger corporation like Apple or Microsoft or whoever it is. And not only introduce me to the person I'm trying to meet, but explain to them like, hey, I want to introduce you to X, Y, and Z. They're doing some really cool stuff in your space, and I'll provide the examples. And then have them say like, this person is really legit. You should be working with them. They're amazing. And I found that I get really massive contracts from that strategy. Now you may be able to, you may be thinking, Well, Neil, this is really hard for me to replicate. And it doesn't matter if you use AI or not, it is going to be hard for you to replicate. The things that work out really well is I've built these relationships over a long time. So you got to put in the work there. I built up a personal brand, which makes it easier to convert and close people. So you got to put in the effort and put in the time and years there. And my corporation has amazing track record, a lot of awards. We're scaled to size. And have tons of case studies that we can end up showing. So all those combined with the correct introduction from the other person who they don't look at as just like an employee or someone beneath them, but someone at their same level, like a peer that they respect, that works really well. It's like if someone has a hot flying startup that is now worth $4 billion and they look up to someone and that someone does an intro about me to them, There's a really good chance I can close a big deal.
Yeah. You know what I'll say? So, 2 things here. One, impact, confidence, ease. So, those of you saying, okay, I can't do that right now because to Neil's point, from an ease standpoint, Neil can easily reach out to people because he already has this network and that makes it easier because he's built this reputation over time. So, you're not going to be able to do that overnight. But from a confidence standpoint, I'm fairly confident on a scale of 1 to 10 that this works literally because Neil's talking about it right now. Impact can be huge. So let's just rate this real quick, 1 through 10 on all of these, right? So impact, I mean, to me, this is like a 9 out of 10 at the least. Um, confidence, probably like a 10 out of 10 because you're proving that it works. But ease is where it's easier for Neil right now because he's put in the work over, God, what, 24 years of doing business, right?
Yes.
And so don't— like, you might as well start now. It's okay to say, like, even if you start now, you compound over the next 5, 10 years or so, and you start to execute this over and over, You're going to do a lot better than your competitors.
Yeah. And to clarify on one thing, actually 2 things, ease, I would probably rate this as like a 1 out of 10 or a 2 out of 10 max. I think it just takes a lot of years before it produces ROI. And the second thing that I want to clarify here is when I ask for the introduction, I don't just hit up people asking for the introduction. So let's say if it's a friend, let's say it's Eric, and let's say he knows the CMO at Apple, I'm making this up, for example. I would say, hey Eric, I see that you're connected with so-and-so, they're the CMO of Apple. Do you know them well? What's your relationship? And then I start diving in with Eric and I'll just ask some questions. I'm like, hey, do they respect you? Do you know them well? What do they think about you, your marketing strategies, all the stuff you put out? This will give me a really good understanding that if Eric does the introduction, that it will help or it'll hurt or do nothing. Because if it's going to do nothing, I don't care for Eric to make the introduction. If it's going to hurt, of course I don't want it. Because some people are connected with others on LinkedIn and they don't know them. They're like, oh, they used to like me, they don't like me now. And you'll be shocked, that does happen. But the key is like, if Eric says like, oh, I know him really well, I'm like, oh, how well? Do you talk to him often? And Eric may say, yeah, I talk to him once every month. And I've helped them actually solve some marketing problems and I've given them 4 or 5 hires and they really respect me in these areas. So yeah, I have an amazing relationship with them. I was like, cool, can you do this intro? This is what I did. Here's a public case study of Après competitors. They spent $4 million with us and look what— I wouldn't probably disclose the price, but I'm like, they spent a lot of money with us. Here were the results. I want to do this for Apple. And I have an opening because we stopped working with them. I can now work with Apple. Let me know if they're interested or I'm just going to go back to the other company and see if they're interested. And I'm not trying to play the companies off of each other. I'm just letting them know, I either can do one or the other and maybe neither any of the companies want to work with me in the future and that's fine as well. So I'm not going in being like, if you don't sign up with me, I'm going to go to your competitor. I'm not trying to do that. It's just more so, If our contract's large enough, sometimes people make you sign an NDA or no, not NDA. What is it? Non-compete, not non-compete, but you can't work with competitors. I forgot what the exact term is, but in the contract it says like, you can't work with X, Y, and Z competitors. But yeah, it's just like, that's an example of a strategy that I found that to work well.
I think it's good. So a couple of things here, I have 2 things. So one, those of you listening to this right now, take this segment, take the transcript from this. If you're watching this on YouTube, go to youtube2transcript.com and take the strategies that we're talking about. sound interesting to you and figure out how you can identify these. What Neil's talking about right now, if you have some of these relationships, what's my play? Look at all my LinkedIn connections. Here's the export of my LinkedIn. How can you do this? Oftentimes, Neil, because my OpenClaw has all my LinkedIn connections in it, I'm just like, hey, who do I know here? Hey, I'm going to Mexico over here. I'm going to Mexico City. Who do you know over here that's an entrepreneur? Who should I— because I'm pretty good at setting up dinners and setting up these small events. And so, I'll do that, right? So, you got to play to your strengths. Here's a question, Neil. Theoretically, this can never stop because you're always going to want to grow more. So when are you going to stop doing this?
I need to slow down because I think it's affecting my family in a negative way. Like, my kids are starting to cry being like, I'm not involved. It's not even spending time because when I'm with them, I'm really like— I bought a crazy Formula 1 big Technic set. My kid likes big cars, or my son does. He's not old enough to do it. So I'm sitting there doing, uh, there's 9 or 10 packs in there. I'm doing a pack a day and I do as much as I can with him and then I finish the rest of the pack on my own.
Technics is like a Lego or something?
Uh-huh. It's like an advent— it's sold by Lego. So it was in a Lego store.
Okay. Yeah.
Yeah. It was like a big Formula 1 car. Um, but like he gets burned out after a while cuz he's 4 and it's meant for 18-year-olds. So like I'll, I'll help him push piece or he helps me push pieces and then when he's like done, I'll finish the rest of the pack. And then every day I'll do one. But it's not the quality time that they're looking for. Like, I even do dinner dates with my kids, just like me and the kids by myself. Gives wife a break. She doesn't ask for the break, but I want one-on-one time as well.
Yeah.
But the kids are more upset that I'm not there for their school moments, that even though my parents weren't always there because they had to work and I understood it, I have the optionality to be there for for my kids financially. My parents didn't. My parents, it was more so if they weren't there, it would be less money for food on the table. So they had no choice. I have the option. And that's where my kids are getting hurt. Because when I tell them, oh, but I'm working hard so you can have a pretty room and all this kind of stuff, my kids keep saying, I don't need the room. I don't need the toys. I don't need any of this. I want my dad.
So they're not wrong. Correct. But here's also like, One, I'll say they're not wrong, but that's on you at the end of the day to decide what you want to do. But also, I think about my past as you're saying that. I'm like, did I really care how often my parents were there? Not really.
Me neither.
Not really. Maybe the only time I cared was when they— like my drumline competition my senior year because we were actually good that year. That's the only time I cared. But other than that, I don't really care because I enjoyed the work and the work at that time was playing video games. So, that was the work for me and you were already working. So yeah, but we're all wired differently. So anyway, that was a real question. You can play this game infinitely if you want to. So anyway, that's a good place to end it. We will see you all tomorrow. Don't forget to rate, review, subscribe. Bye.
