Your Google Analytics Is Lying to You, 57% of Your Traffic Isn't Real

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. So 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 change 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, 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 going to start to really destroy the job market. I think there's going to be job displacement, but I think net-net, there's going to 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 going to create more jobs long-term than destroy jobs.

We see it in marketing too. We believe AI is actually going to 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, you know, 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—

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, 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 it's 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— 10 years. 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, I was 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. Now I'm finding so many people to hit me up asking me questions about how 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, you know, 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, uh, and he hired him. He's been there probably for like 8 to 10 years now. Matt runs his engineering. He's like, dude, you know, 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, uh, 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? And from 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. And which is interesting to me because it's like, it's 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?

Yeah. Yeah.

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 with, for just kind of 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? It's constantly more, more, more, more, more, right? And so going back to the initial topic that we had where the guy wants everything, 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.

Good chunk, 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, right?

And so actually, by the way, go ahead, finish your thought.

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, it's 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, like if it's the status quo, like it's 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.

That's— we'll see. No, I think it is a 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 AnswerThePublic and Ubersuggest, we should 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.

Yeah.

Unless you have someone focus on, hey, I'm going to focus on product marketing and just, you know, create material around these new features and videos and social posts and, you know, do SEO campaigns around it and paid advertising campaigns. But it's becoming overwhelming because of how fast people are moving within organizations now.

Just a quick break. If you want to run LinkedIn personalized ads and landing pages, you can't do it on LinkedIn ads right now, but you can do it with Karrot. It's www.karrot.ai. Again, karrot.ai, and we'll see you on the other side. Yep. I'm going to show you something real quick. Like, if— 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. And basically what we're having people do now is create artifacts, right? And what I mean by an artifact, if you see this on the screen right now, is we're having people create HTML documents. So 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 kind of 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, you know, 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. Um, but this is how we're talking now because it's like instead of having to read the entire transcript, it's like that's— it's hard to parse that information, right? 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.

So yeah, no, I'm with you. Communication is 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. I remember Eric used to make me take surveys or quizzes. I don't know what they were called.

Personality tests?

Personality tests. There you go. And he would make friends take them and other entrepreneurs that he knows, and it would compare—

I 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 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 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 could 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 don't remember. 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, the reason why, 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? But the thing is, when I looked 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 kind of 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, you know, like 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. 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. 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. Uh, although, I mean, that data is like not that bad. So, okay, let me pull up this Paul Graham post. So Paul Graham is the founder of Y Combinator. So Neil, I want to 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 an 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 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 them. 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 going to 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 that CEO is 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 like 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 Not— 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. 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, like 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 he, 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? ClickFlow 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 Google bot or, you know, 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. So now you're really optimizing for 3 things: the platform bots, agents, and the third one is humans.

Anyway, thank you all for listening. Hope you enjoyed it. Please don't forget to rate, view, and subscribe, and we will see you tomorrow.