Will Google's SGE Ruin Traffic? and Elon Musk's AI Chatbot Grok (AI Corner Dec 2023 Week 3)

All right, so we are going to talk about AI Corner. So we've got a lot of cool bells and whistles here to talk about. And you know, I'm going to start off with this one over here. So Neil, have you seen this Wall Street—

No.

You definitely haven't, but I'm going to act like you did see it. So there's this Wall Street Journal article that came out and the headline here says, news publishers see Google's AI search tool as a traffic-destroying nightmare. Okay, so let me share my screen over here. And you know how like publishers always complain. And I think it's good for everyone to understand that there's always been a fragile kind of like agreement between Google and publishers, right? And a publisher is someone that might be like a Wall Street Journal, they're publishing content or a Business Insider. Now, that tentative agreement has always been, hey, Google will send you traffic and the publishers, the agreement from their side is we will supply you content. And now, so what's happening now with the search generative experience is that Google's scraping a lot of the web and just giving you an AI answer, right? Which by default means that it's taking away clicks Which means it's taking away traffic, which means that Google is taking away revenue from a lot of these publishers. Okay? So when you look at The Atlantic, which is a publisher that's been around since 166 years, Steve Jobs' wife, you know, she ended up buying it. But they think that 75% of the time, the AI-powered search would likely provide a full answer to a user's query and The Atlantic site would miss out on traffic it otherwise would have gotten. So that's one number over there. Now, the other thing here is publishers have seen enough to estimate that they will lose between 20% and 40% of their Google-generated traffic if anything resembling recent iterations roll out widely. Google has said it's giving priority to sending traffic to publishers, which of course it's going to say, right? The last thing I'm going to call out here before we react to this, Neil, is this is the Liz Reid, Google Vice President. So here's what she said. She said, basically, any attempts to estimate the traffic impact of our SGE experiment are entirely speculative at this stage. And we continue— as we continue to rapidly evolve the user experience and design, including how links are displayed, and we closely monitor internal data from our tests, which is a typical Google answer when it comes to this stuff. So let's talk about this.

I agree with that because, you know, I've talked about this in the past and I'm actually pulling up the data right now and I'll share it on my screen once I have it. But everyone's like, oh, SEO's dead, SGE. Well, first off, Google doesn't even know what they're going to end up doing with SGE and all the changes. Like, it's so early. It's not something that's going to end up rolling out in a matter of, I don't know, like weeks. It's going to be—

If they screw publishers over, they don't have a business.

Yes, and you got it right because one, they need it for content. Without that, they really are screwed. And 2, SEO, more so Google AdSense, really drives a lot of revenue. So check this out. I'm going to share my screen right now. I just got the stats. People talk about SEO being dead or they don't want to drive traffic to publishing websites. All right, so Google made $32.78 billion in revenue from their partner network, AKA AdSense. And this is in 2022. And if you had to guess how those sites get their traffic, it pretty much is SEO. If you're making money from AdSense, AdSense is people buying traffic from Google Ads. You yourself cannot buy ads on Google and then send it to AdSense and make it profitable. You will lose money. Some people try to do it with Yahoo. Some people try to do it with Bing or Facebook and then send it to Google AdSense. Those days are long gone. It's very, very, very few people who are able to do that profitably. Those businesses don't last long and it's hard to do it at scale. You know, it's hard to arbitrage from one ad network to another. But on the flip side, the way you get your traffic is organic social and organic SEO. Organic social doesn't give you consistent traffic. Organic SEO does. If they just killed SEO, they're going to kill a lot of their revenue. Now, am I saying that Google isn't willing to crush $30-something million in revenue? I'm not saying that. I'm saying they're not willing to crush it and, you know, bank on AI until they know how to make up that revenue difference through something else. And they'll run a ton of experiments, which is the Google way, until they figure out what produces a better ROI. And they're very, very careful to do it with a fraction of a percent of their traffic or 1% does good. Then they'll start opening up to maybe 5% or a region and then 10%, then 20%, then 50%, and then boom, they'll just roll it out once they know it's going to make them a killing. They're not going to do something that's just going to crush their stock price. They're really smart and sophisticated.

I just think, look, the people that are complaining like Barry Diller, who's Expedia, Zillow, and all that, all the publishers that are complaining, The majority of their traffic comes from organic. And so yes, what happens is you start to complain because when the golden goose starts to like disappear where you feel like it's at risk, you start to panic, right? So what is the counter to this? The counter is to take an omnichannel approach. We talked about podcasting, YouTubing, all these other channels becoming bigger and bigger. It's time to diversify and not rely so much on Google. And Google's been good. Like Neil and I come from Google, right? Like we come from SEO, like we love it. But we also know that we can't rely on it because then otherwise we'd be panicking right now. Yep.

Speaking of AI, another thing that we've been testing out a lot is the different platforms. So ChatGPT, of course, Bard, of course. Eric, have you played with Grok a lot?

Well, I need to wait one more day because I was too cheap to— so I, anyway, I didn't want to pay. No, no, the iOS, there's a 30% tax on it, right? Because Elon doesn't want to, so he makes the users take it. So I'm like, oh, I'm going to pay for Grok. But it's like, oh no, you have to upgrade to the iOS version. I was like, I don't want to pay for it. So the support's like, you need to wait until the 20th. And so tomorrow I can finally pay for it.

You can just upgrade on your computer and you don't have to pay the 30%.

It doesn't let me because my thing expires. My iOS expires on the 20th and then I can finally say, okay, I'm going to sign up for the web and then I can pay $16 instead of $22.

Oh, okay. That's not too bad then. So what we found with Grok is, You know, ChatGPT by default ends up posting, creating shorter content than Grok. So we kind of like that. Bard actually creates more in-depth content than even Grok. But from length perspective, we think Bard does it best, then Grok, then ChatGPT. What we're also seeing is Grok has a lot less rules. Like if you ask Grok to use, you know, profane, or if you ask it to be vulgar when it's creating something, it'll be vulgar. If you ask ChatGPT to, It will not be vulgar at all. You know, it'll be blocked out. Now, what we found is Grok also creates content with more emotions. It does a really good job at storytelling. Bard does too. ChatGPT is not bad as well. But with ChatGPT, we found that to get some of the things that you're getting from Grok and Bard, you can get. but you have to be more descriptive with your prompts, right? While with Grok and Bard, you don't have to be as descriptive with your prompts and a lot of times you can get at least marketing content that's sufficient.

Yep. Yeah, I'm keen to try it out. I think there's a lot you can do with it. I think they have a very unique dataset, so X does, and I think they're going to be able to build a lot of interesting products from in the future. So I'm excited to see what happens there. Next thing over here is, We wanted to talk about more identity layering ideas. So I've talked about how you can use tools like retention.com or customer.ai or Wormly. There's a lot of these out there where you can deanonymize the IP addresses that visit your site. But for example, when I look at a report that come— I get a daily export from retention.com and I will see the director of marketing from Shopify visit. I will also get their LinkedIn. I'll get their email address as well. I'll also get their estimated revenue, number of employees. I'll also see when they first visited, the first page they visited, and the other pages that they visited. And so now I have a lot of contextual data I can go after. And so let's say Neil visits my website. I might say, okay, he visited a page on SEO services. I might say, hey Neil, thanks for visiting our site in regards to SEO. Have you looked into programmatic SEO or something like that, right? Or have you looked into this content over here? Like this might be helpful to you. My point of saying all this is that With this data that you have coming in, you can leverage the power of AI to find the right people, add them on LinkedIn, send them a message that's contextual based on whatever it is that they were interested in, what they visited. And then you can send them more content that's relevant to that and you can build a relationship there, right? You can add them to your email list, you can retarget these people. But the other thing too is most people don't think about identity layering from, hey, what if you're looking to recruit people? There's plenty of people you're going to see on your list that are visiting your website, right? What if you're looking to do like a— like you're looking to acquire more companies? You might have people that might be potential acquirees visiting your website. So there's just a lot you can do from a recruiting, a marketing, sales standpoint, an acquisition standpoint with this stuff and layering on AI. You can use clay.com to do it. I'm just throwing a lot at you right now. I'm just saying, go look into this stuff. And I think this is a lot that we're going to be focusing on in 2024.

Cool. So what's the last one you want to end up covering on AI?

Well, don't— you know what? I want to talk about not overloading on AI tools, right? Like, let's start with that. I think you have a point of view here.

So with this right now, everyone tells me about their latest and greatest tool. It used to be like Jasper and they're like, oh, you got to focus on ChatGPT. You got to focus on this tool. They're doing some crazy stuff with their API. And the way I look at it is there's too many tools coming out. Too many of them are features, they're not actually solving a big problem. They're just doing one little piece of the puzzle. And what my belief is, is focus on most of the major platforms, you know, Bard, ChatGPT. If you want to throw in a third, maybe Grok, but it's not necessarily needed. But focus on ChatGPT and Bard because they can do most things for you. And then from there, wait, no joke, 6 months to a year, and then see what's out there that still sticks. Because most things will just go away and people are like, oh, you know, you can do this with your podcasting or images. Well, eventually ChatGPT and Bard will do all that kind of stuff for you anyway. So why would you need to go pay for all these tools? And ChatGPT and Bard are probably going to do a better job. I'm not saying that because, you know, they're the smartest people in the room or anything like that. That's not the reason why. I'm saying it because they have the most capital, right? When you're spending $2 billion, $3 billion, $5 billion on whatever it is to make these tools amazing, you bet that money is going to go a really long way and they can afford to pay more for better engineers, which is required for a lot of this stuff, than the competition. Even if you're venture funded and you have $500 million, it's a drop in the bucket compared to what some of these other guys are spending.

No joke.

That's why I think you should just mainly stick with ChatGPT and Bard and not worry about the 100 other tools that are out there. I think it gets to be so bad and overloaded that it's a time suck and you spend so much time working on these tools that are features, you got to then spend a lot of time to clean up the mess that they're creating. And then you're like, oh wait, I could have just done this manually faster.

Yep. Look, I also think a lot of the AI tools out there being built are just crap. They're not defensible and they're going to be commoditized very quickly. And what's going to end up happening is— Jasper.

Jasper is a prime example. That company will be worth very little to nothing. Mark my words. Maybe people will disagree with me, but you don't need it to create content. You can just create it with ChatGPT or Bard. Skip the paying money.

Yeah, look, I like the guys at Jasper a lot. I do think it's a tough situation. I remember when they first, when ChatGPT first came out, I believe the founder, the founder Dave, he reached out to Sam Altman and he's like, hey, like, you know, hey, I think it was something along the lines of, hey, like it would've been nice to have like a heads up around this or something like that, right? But when I look at, not even Jasper, I think looking at other tools too, it's like, to me, they're basically helping you prompt engineer a little bit and it's nothing proprietary. And what is proprietary at the end of the day, one, yes, focus on your business, focus on what's actually going to drive high leverage. But if you want to play in this AI space, if you really want to have a tool, I think the people that are going to win at the end of the day are the people that have their own proprietary datasets. And the ones that like when you layer on your own dataset in addition to like a Bard or a ChatGPT, for example, that's where you're going to get amazing insights that are going to be able to move your business forward, right? That's where I think the money is and that's where I think you can make the biggest impact because I hear about a lot of people right now. It's like, hey, what I've done, I produced 4,000 more pieces of content or whatever. And it's like, dude, you use like 20 different tools and like all 4,000 pieces of content that you produce are all garbage. And so like, sure, you're producing it, but at the end of the day, what results are you getting? And more importantly, how much time are you wasting doing this?

Dude, I'm with you. Well, that's it for today's episode. Make sure you rate, review this podcast, 5 stars. We really appreciate it, and we look forward to talking to you and seeing you tomorrow.

And check out this next video over here. Goodbye.