Ep 236: How SurferSEO AI Is Changing The Game Of AI Generated Content Massively with Michal Suski

Jaryd Krause:

Imagine a world where AI content doesn't just create every piece of content for your website, but it does it in one click and outranks your competitors. Hi, I'm Jaryd Krause. I'm the host of the Buying Online Businesses Podcast. And today, I'm speaking with Michal Suski, who is the co-founder of Surfer SEO. He is a digital marketing and website quality enthusiast. And as an SEO genius, he has a data-driven approach to optimization. He has also trained hundreds of people on on-page optimization, SEO content, and technical optimization, and now at Surfer SEO, he heads up innovation for the tool.

In this podcast episode, Michal and I speak about how Surfer SEO has evolved and how it worked when it first came out many years ago. And how it's working today, the linear process that you can use to create content through the tool, and why it's valuable. But most importantly, we dive into so much in this episode about how Surfer SEO introduced Surfer AI, which creates content super cheaply.

But not just that. There's a huge surprise that is an absolute game changer in the AI content generation space that I actually go away and suggest in the podcast episode. I bring it up as a thought process and a suggestion. And Michal just blew his absolute mind, almost wanting to stop the podcast and start writing things down, taking notes, and going down different paths. His mind was going insane.

We also talk about how AI will change websites and our space online for the better. And we discuss the common fears that most of you guys are having around AI content and the future of content websites. And we showcase why content websites are absolutely not going anywhere and are only going to get better.

This is such a valuable episode. I'm just buzzing, thinking about you listening to it. So I have no doubt you're going to absolutely love it. And please do us a massive favor and share this podcast episode around, because it is mind blowing. See you on the inside.

Do you have a website you might want to sell either now or in the future? We have a hungry list of cashed up and trained up buyers that want to buy your content website. If you have a site making $300 per month and want to sell it, head to buyingonlinebusinesses.co/sellyourbusiness or email us at [email protected] because we will likely have a buyer. Details are in the description.

Welcome to the podcast. Am I pronouncing your name correctly, Michal?

Michal Suski:

Yeah. Pretty cool. Pretty close.

Jaryd Krause:

Awesome, awesome. What is the correct pronunciation?

Michal Suski:

In Polish, it’s mixaw. Like me doing something, how to do something.

Jaryd Krause:

I like that.

Michal Suski:

If you connect these two, it's perfect.

Jaryd Krause:

Yeah, mixaw. It's very similar to hello in Chinese.

Michal Suski:

Really?

Jaryd Krause:

Yeah. I think ni hao is hello; I might be incorrect, but nǐ with an N.

Michal Suski:

Nice.

Jaryd Krause:

Yeah, I think that's right. Guys, correct me if you're listening.

Michal Suski:

Googling right now Ni hao, yeah, it's pretty close. It is. It's N instead of M, but there's no big difference.

Jaryd Krause:

It's cool.

Michal Suski:

Thank you.

Jaryd Krause:

So welcome back to the podcast.

Michal Suski:

Happy to be here. Thanks for having me again.

Jaryd Krause:

You're welcome. So episode 62, for everybody listening and yourself, was when you were on last, and that was called How to Boost Your Website's Rankings with Surfer SEO. And so we're up to late 230 now recording this one. So we've done a fair bit. It seems like it's been a fair few years, and I didn't realize it had been that long. I'm looking forward to digging into some questions here.

Michal Suski:

Do you record weekly?

Jaryd Krause:

Sorry?

Michal Suski:

Do you record weekly? Is it weekly?

Jaryd Krause:

Yeah, weekly.

Michal Suski:

Yeah. All right. 50 podcasts a year. Nice. So it's been a few years. Three years, man. Time flies.

Jaryd Krause:

Yeah, just over three years. And so there've been so many changes with Surfer since then. I think you just kind of got the ball moving with Surfer four-ish years ago, right?

Michal Suski:

Yeah. It was 2017 when we established an informal connection to start something.

Jaryd Krause:

Surfer SEO, known for helping people create great content. You've got a Keyword Research tool, and you've got a Content Editor tool. Tell us, when somebody's looking to work with you, what's the number one thing that they're trying to solve with their blogs?

Michal Suski:

Yeah. Basically, to spend their time well. Because when you are spending your time on writing and it takes four hours, five hours, whatever, to craft something nice, then it would be a waste of time if you didn't optimize it for SEO, unless you have another stream of visitors, of course, like social, email, and stuff. So if you are writing for these channels, then that's fine. But if you are looking to get some extra revenue from search, then it would be a total waste of time if you skipped the optimized writing process, which is basically what we do.

Jaryd Krause:

Well, let's talk about the evolution of Surfer, then. Because there are so many things you've added on, and we're going to eventually move into AI. And it's what I wanted to bring you on to talk about. It's probably going to take up a large portion of the call. So how did Surfer start, I guess, and then what were the iterations of the products that you brought on, and why did you bring on those different services and products? Because when you started, I think when we first chatted, you were mainly helping people that had already created pieces of content, like plugging Surfer in and finding LSI keywords.

Michal Suski:

Yeah. Analyze the content that was already there, right? You are right. And it all started with in-depth analysis—a ton of ranking factors that were analyzed for a correlation with the SERP to tell you, like, Okay, change this or change that, add more content, reduce some headings, whatever. But then people started looking for something simplified to scale up. Instead of analyzing 200 or 300 factors, they wanted to take the 20 most important, implement them, and move on to the next page. So from the in-depth analysis of the charts and correlations, we implemented a so-called audit feature that was a simplified version of that big chart that was showing all these 200 factors.

But then, like you noticed, they were discovering keywords to include and stuff, but then they had to implement these suggestions. So work with the text, basically. And one day, I was talking to one of the users, and he just sent me a photo of his writer who was really writing down keywords on the paper and crossing them with a pen, which was mind-blowing.

Literally, she took the audit recommendations and wrote them with a pen on paper. And while she was optimizing, she just crossed them with a pen to understand, Okay, I got it. I got it. I got it. Boom, boom, boom. And all the way down. So we were like, Come on. It should be easy to read the content while you are working on it to dynamically show the suggestions.

Jaryd Krause:

Correct, yeah.

Michal Suski:

And that was the moment when we introduced the Content Editor, the tool that you mentioned, the one where you literally start over with writing or fetch your existing article and work on it. So not only were the suggestions there, but you could also implement them along the way. You are writing, and you see what's going on. You see how your changes to the content affect the overall score and the optimization stuff.

And it was like, All right, now they know how to write. Now they know how to optimize. But they still don't know what the next best topic is to cover. And we've fluently transitioned to keyword clustering tools that are revealing the whole set of ideas that you should consider if you want to build authority on a given topic.

Jaryd Krause:

To build topical authority?

Michal Suski:

Yeah. But that was before the topical authority term was even there. It was 2019, I think. It wasn't so popular back then. Yet we had a tool for that. But that was a pretty cool tool set. So you know what to do with existing pages and how to implement the changes. You can either write new content or optimize existing content through audits and stuff, and then you know what to write about.

But there was another struggle: first maintaining the content that you had already created. So not only should you periodically revisit your audits because SERPs are volatile, but things are changing in Google every quarter at least. I would say it's a good idea to review your best-performing posts to just make sure they are still up to date with the entities, with the LSI keywords, and with the optimization rules that Google is promoting right now.

So we introduced the Grow Flow, which was like a sort of autopilot on your most important pages to tell you what the missing keywords were every week or so. It wasn't a big success, though, and we still work on it. But moving forward, there will be a lot of iterations, really. On the next episode of your podcast, we will spend 30 minutes of me talking about what has changed.

Jaryd Krause:

That’s funny.

Michal Suski:

Yeah, kidding. So there was this thing of research, the clustering thing. It produced about 100 ideas. 200 topics to cover, 300 topics to cover—in these bigger, more competitive niches, it was just the tip of the iceberg, right? And there was this problem of how to scale up my writing process. At some point, Jasper showed up, right? It was like 2021, I guess—yes, it was—when we established a partnership with them to help Surfer users produce their content quicker.

But they were still wanting to have more and more articles without the hassle of really prompting, without the hassle of typing something, then receiving something from Jasper, then typing something else, and then receiving. They were like, “We want to automate. We are SEOs. We want to automate everything possible. And we don't want to deal with too many people at the company.”

Jaryd Krause:

We don't want too many hands to automate the content because it can just break it.

Michal Suski:

Yeah. And we are kind of fluently moving to AI, I guess, because this is something that moved the needle for us. It was like five or six days ago when we released it, but the traction is incredible. And it feels like we hit the nail on the head with the demand for it and the solution itself that works for the people, really. Really looking forward to seeing what happens in the coming months. Yeah. So that's the short story.

Jaryd Krause:

Yeah. That's the short story of seven years, in which we haven't touched on AI yet, and we will. So initially, it was just a tool where people could go away and do their own keyword research. They would have to not yet use a tool, do their own keyword research outside of it, create the piece of content, plug it into Surfer, get the optimizations, and then go away and publish it on their site.

And then you brought in the Content Editor, helping people edit the content based on what the current environment is looking like in Google at the time and what SEO would help that piece of content rank higher and be better with those SEO tweaks. And then you brought in the Keyword Research tool, helping people build out topical authority and clusters.

And as that's evolved, there's been a lot in between. I know that. And people in our community use Surfer and love it. And that's another question I'd like to ask you later down the track about single website owners versus agencies and the different types of users you have for Surfer.

But let's jump into AI because people want to make it so much easier to create content. And Jasper came in, and now Jasper had to get better because of ChatGPT. And it's only getting better and better. And I'm sure you can perceive that happening with Surfer as well as as AI evolves. So is Surfer AI an add-on and how does it work?

Michal Suski:

Yes. So basically, you just get credit for the AI. And for that credit, in exchange, you receive an SEO-optimized article that will be long enough to run, that will have these keywords that should be there, these headings that should be there, topics touched, and a coherent structure. There's a lot to it.

To all of you who are familiar with AI from what's under the hood, we use up on average 400,000 tokens on creating one article. And for those who are not that familiar, it's more or less 300,000 words either read or produced by AI in the various steps of the process to receive this finished article that is around 3,000 words long. So it's like, yeah, 100 words in prompts equal one word of output. So there's a lot of information. There's a lot of information in the pipeline. That's why it's different, basically.

Because, well, there was this one thing that the CEO Lucjan told me, “Let's make it the best and let's not look at the cost, really. Don't try to optimize the cost, please.” Because we had a moment in time, in the past, we tried to implement GPT2-XL, the extra-large model that was fine-tuned on every keyword and stuff. We built pipelines. But it wasn't that good. I mean, it wasn't even close. But it was almost free, right? So we had this approach.

But now we have a different approach, which is to read as much as you can, understand as many topics around the keyword as possible, and spend a lot of tokens to teach the AI what the hell is going on. Is it up to date? Is it factual? I mean, factual, depending on whether the setup is factual. But there's not much we can do about it, right?

So do as much as you can to understand the topic and produce the article that will be the best. And don't look at the profit margin. Don't look at the cost. And this is where we ended with a premium, of course, kind of feature. You can get cheaper articles. And here, it's like 29 bucks. It's sort of the highest tier of AI articles. However, we are not taking a massive cut. We are spending a lot of this money to give it value. So that's the difference, I guess.

Jaryd Krause:

Yeah, cool. So it costs $29 for the user to use it. Is that what you said? It costs $29?

Michal Suski:

Yes, it costs $29, but it's per article.

Jaryd Krause:

Yes. So $29 per article is okay. Yeah, yeah. All right. And so people are like, Wow, if I can get this done and it's amazing, why would I not just spend $29 on a piece of content that is spending however much they're using with an agency?

So how does it work for somebody that's listening and is thinking, Great, Surfer SEO is going to basically do what AI does in, I don't know, I'm generalizing here, go away and create this piece of content? But what are they doing? Like somebody who's a user, I want to get a piece of content created. Do they go and get the keyword research done?

Michal Suski:

Just one click. Just one click, and the article will be there. That's the difference between experience with ChatGPT, Bard, or any other AI, OpenAI Playground, or whatever. When you produce content on your own with AI, it's a matter of you passing the information to AI to avoid the data cutoff because the data set is trained on information from 2021 and before. So it won't be that factual on the recent stuff, and well, things change. So you have to feed it first.

So it will be a matter of prompting over and over by hand, like, “Hey, read this piece of content and tell me what are the topics mentioned here.” And you wait ten seconds, and you receive these topics. Then you take another paragraph, another section, and another competitor. And after 30 minutes, or if you automate it, maybe a little less, you can receive a set of topics to mention within the article.

Then you have to figure out the headings the same way, and then you have to figure out the keywords. You can use Surfer, you can use other tools, you can use your own, like Python scripts, whatever. But basically, it's back and forth talking, “Do this,” okay, I received, then “Do that,” then I received over and over.

Jaryd Krause:

So the prompt engineering is sort of built into the tool.

Michal Suski:

Yes, exactly. We did the whole prompt engineering thing. This is where these 400,000 tokens are. We do prompt engineering to understand everything, compile it, and then create a sort of knowledge base about what should be mentioned there. And then the AI utilizes the knowledge that it itself researched. It's just like if you were a regular writer who was asked, “Hey, do research on a topic and then write it.” This is what we are doing. We are using AI to do the research. The research is done, the information is obtained, and then the information is used to produce the article.

If you are writing on your own without prompting, you just receive bits of words. If you have a longer prompt, you can receive more content, especially when using GPT-4 and browsing plugins. There's a plugin called Webpilot. You can use GPT-4 with a chat that will be able to browse the web, so you can ask it to go to the website and do something with the content, and it speeds up the process.

Yet, if you want to really scale up, you will start to notice you are losing time. And if you can, if you value your time highly enough, there is no other way than using solutions like Surfer’s. However, if you are working on your own and you are just keeping up, trying to get there, and not necessarily spending a ton at the very beginning, I would say that GPT-4 with browsing plugins will be better for you just to trim the frustration of initially spending a lot.

Jaryd Krause:

Yeah, I totally get that. Now, where people are going to think is, “Cool. With GPT, I need to go away and make sure that my content when I produce it, it's not sort of getting detected by AI generators and it's really damn good. And it looks like it's human generated.” Even though Google says, “We don't care.”

Michal Suski:

Google cared?

Jaryd Krause:

Google says, “We don't care how it's produced as long as it's good.” What's the thought process around somebody using Surfer SEO AI going, “Right. Here's my $29 per piece,” and then publishing it? Do people have fears around... If you know what I’m saying.

Michal Suski:

I did a lot of analysis on it. Even with the one that is probably the best—I mean, maybe not the best, but the most spoken—is Originality AI. And even with their recent changes to the algorithm, we tested it. And well, it was 50-50. I can tell you that we analyzed some 200 samples from content editors of our users who typed the content in. And these content editors were old—before the AI, before Jasper, and stuff. So we have everything in the database.

So we took these. There's no way they were AI, right? And we took them, and 50% of them were considered very likely to be AI, which is nonsense, right? Even the Bible. The Genesis from the Bible was like 99% AI. This is something that comes from the community. It was this guy, Alexander Lubinkovich, I guess. He shared a screenshot. “Hey, take a look. Genesis is 99% AI,” which is another ridiculously weird example.

And another one, Dennis Hamming, was a little bit controversial, but he basically took the GPT-4 output, straight off GPT-4, swapped some commas, and basically did nothing to the content. And from AI to human in just a little change of replacing commas with something, with whitespace or something, I can't remember. But it was really not a significant change to the content itself.

So there is really only one use case for these tools. If you order content from a human writer and want to know you are not being cheated, you can use that product. But if you want to produce AI content with the idea in mind that I will go straight to the AI detectors and put my AI content in there to see if it is AI or not, it will be hit or miss, 50-50. There's really no point in doing it. So, yeah, there is one application, like I said, for these tools. The second is educational, for the teachers and stuff. But for SEO, just keep the quality high, and you'll be good.

Jaryd Krause:

Keep the quality of the content high.

Michal Suski:

I even got a case study, man. Let me quickly bring it up because I almost forgot. We had a website ranking pretty well in the marketing niche, sort of the marketing communication niche.

Jaryd Krause:

Digital marketing.

Michal Suski:

Yeah, digital marketing and building businesses are basically like a hustler's website, right? So we had this website. It was kind of topically authoritative. It was getting traction—a lot of traction. And it was produced by AI without the best quality.

Jaryd Krause:

All of the content was produced by AI?

Michal Suski:

All of the content. All of the content was produced by AI with a few human edits. But basically, an AI website of 3/10 quality is enough to pass, right? And it was growing. And boom, it went down. A subtle cutoff.

Jaryd Krause:

I’ve seen this happen multiple times.

Michal Suski:

It went down like crazy. And we were like, “Okay. There's nothing we can do about it. That's bad. It's a pity, but whatever.” And some time passed, and we started to work on Surfer AI at that point. And it was like, “Hey, isn't it the best playground we can have to replace these poor AI articles with our AI articles to see if it moves the needle?” And you know what? It moved the needle. It moved the needle. The site recovered. We replaced, like, over 100 articles. And some of these articles, compared to six months ago.

Jaryd Krause:

So you updated the content that was no longer ranking? Is that what you did?

Michal Suski:

Yeah, the content that was taken down. So there was an article about getting 1,000 clicks a month. It went to zero, and now it has even more.

Jaryd Krause:

More than 1,000 clicks a month?

Michal Suski:

Let me check. I mean, this website right now generates almost two million impressions and 14,000 clicks. So it's kind of decent.

Jaryd Krause:

Per month?

Michal Suski:

Yeah. It's quite significant for a page that’s AI generated, right? And all of the articles, as I'm scrolling right now through the report from Data Studio, are getting 3,500% more impressions, for example.

Jaryd Krause:

Wow. And so, how long ago did you update those pieces?

Michal Suski:

We started updating in February. The page was taken down late last year. We weren't even talking about the AI at Surfer, and we had these first versions in late February or early March. And we started to replace it with the first versions of Surfer AI, and it helped a ton. We will continue to study, probably by building another topical hub entirely based on Surfer AI, just to expand a little. But definitely, it helps. It helps a ton with recovering a busted website.

Jaryd Krause:

Yeah. So that's a really good case study. It's going to help me with the title of this podcast. And people are going to love hearing this. Now, I'm curious, and this could be an extension of a product, whether you guys have it or not, or an extended service in the future. But maybe you already have it.

Is there the option where you go in and you get one piece of content created, say, on a certain topic, and then you realize, Okay, I need a couple of these other topics because you use this other tool that's within Surfer? Is there a way that you can purchase packages of pieces of content, like AI content, where you can just purchase a cluster or a silo? Have you got that, or is that something else?

Michal Suski:

I'm glad you're asking me. Because right now I'm responsible for innovation at Surfer. And this is one of the ideas we've been exploring. We didn't follow up on it, yet I already have like crude versions, I will tell you. Sit down.

Jaryd Krause:

Because it sounds like you've got the capability already with what you've built out, and you don't have to share what you can't yet.

Michal Suski:

Everyone who is not sitting should sit down.

Jaryd Krause:

And if you’re driving, pull over.

Michal Suski:

Because in terms of the bulk, in terms of the variety, but really, what we are capable of doing, and I did it on a beta test internally, was ordering in bulk the whole topical cluster. But that's not it. It was analyzed by machine learning algorithms to connect each piece to each other. Where is the connection? To figure out the internal link and the placement.

To put the internal link in there. And your only task was to tell me what the pillar page here is. And you selected the pillar page, and the whole structure of the hub and spoke, kind of internal linking, and everything to the middle meeting at the pillar page were there. I'm waiting for a case study with one guy, yet he's kind of slow in delivering.

Jaryd Krause:

Also, guys, when you're listening to this, as you pull this up, Michal, this is coming out a little. I'm dating the podcast here a little bit because I'm going away and having some fun traveling and whatnot. This might even be closer to happening to Surfer, I guess, by the time this comes out. Who knows? But it's cool.

Michal Suski:

When will the podcast come out?

Jaryd Krause:

This will come out in September-ish. So it's about September-ish now when we're publishing this.

Michal Suski:
Oh, you may be a prophet here. There is a high chance you’ll be able to do it by that time.

Jaryd Krause:

Also, guys, wink-wink, nudge-nudge, yes. I am a big fan of Surfer SEO and am an affiliate because they are awesome. So there will be an affiliate link to this. Just a full disclaimer here. But yeah, I'm just fascinated by how this has evolved and is evolving.

Michal Suski:

There's nothing wrong with recommending things that work for people.

Jaryd Krause:

Amazing. Our community loves you.

Michal Suski:

You are helpful.

Jaryd Krause:

We use it, right? Yeah, it's good.

Michal Suski:

Helping your people discover things that help you—there's nothing better, really, than this.

Jaryd Krause:

That’s the best thing you can do for the people. So that's massive, man. To have a pillar piece or pillar page and then create a cluster and internally link and have all that content created, and it all makes sense like that. And being able to sell that as a package, I don't know. It'd be dependent on how many pieces you want in each cluster and the depth.

Michal Suski:

Yeah, exactly. It will be like you are in a grocery store and you pick from the shelf, “I need this, this, this, and that.” You put it in here, select the pillar, or select the pillar you already have. Imagine you are boosting one of your pillar pages and you feel like, “Okay, I've got 20 supportive pages.

It feels like I need another circle to support it, to boost it from behind even more.” So you order them like missing pieces, interlink them together, and mark them. Here's my pillar page. So throw the links at the pillar page. And it will use the right anchors. It will use the right combinations of words. So I can't wait. I can't wait. I hope it's already there, available right now, by September. But I don't know. We'll see. We'll see.

Jaryd Krause:

Surfer AI is available now, at the time of recording. Will be when this goes live too. So, guys, definitely check that out. But I'm sure, like you said, you've got that coming. Is there a thought process in terms of innovation, knowing it's one of your key roles at Surfer, around, All right, we've got this cluster and this pillar page, and all the contents created, just to say it's like 20 pieces of content with internal links and all that sort of stuff? And it's about, say, surfing, and it's about picking a certain surfboard for somebody who is a high performer.

And would the tool, like it does with keyword suggestions, have topical pillar page keyword suggestions or pillar post suggestions? Or could it maybe be in the future where, if you're buying a performance, high performance surfboard, maybe to create the linear journey of somebody who's a beginner, you need something, you need a cluster on how to buy a surfboard for a beginner, and then how to buy a surfboard for an intermediate, and before you get to that? That would be fascinating.

Michal Suski:

Exactly. This is another innovation we are working on. Right now, we have two roads, or basically two projects, open. And one project is, of course, scaling up your writing with AI. But with AI, you can write so many posts that you will run out of ideas. So the clustering tool we've been developing since then is evolving because we managed to cluster the whole internet, which is basically a map of everything. We have the connections between things as Google sees them.

So by September, obviously, but hopefully even right now in June or July, you'll be able to see how your website looks on that world map. Like you are playing civilization or whatever, like heroes. You've got some cities taken over, right? But you've got surrounding cities that are not in your territory. So we will be able to show you the map.

Where are your cities? Where are the cities that you should visit to expand your territory? And which cities are the closest, the easiest to take over, and the most beneficial to building the coherence of the topical authority map in general?

So if you see this visualization, you see, okay, here I am here, here, here, here, here, here, here. And around these topics that I already covered, there are other topics, for example, the best surfer for intermediate because you already have one for beginners, and one for advanced. And the closest untapped opportunity is for intermediates. And we know it because it is the closest bubble based on relationships in the SERPs.

So this is something that will be super helpful to actually understand the topical authority because we have this really scientific approach to it. And this is absolutely fine. And I'm not neglecting it at all, but it's not for the people.

Jaryd Krause:

Yeah. You also want that; what's the real journey? What's the real user's journey in real life? What does it actually look like for them? Not based on what's already out there scientifically on Google, right? What are they actually doing?

Michal Suski:

But that's another story. That's another story. Jaryd Krause:

Because yeah, it's fascinating. You can say, Okay, once you've picked this board, you surf on it for X amount of time, and then you can step up to the next board. But hang on, what's in between those? What are the idiosyncrasies of the struggles they have?

Michal Suski:

Some tricks, some tips, right?

Jaryd Krause:

And what is being presented online may not actually be, “Hey, that's what a surfer is actually going through at that time”?

Michal Suski:

I hope I can implement it before it goes live. This is a really cool idea. I didn't think about it. I didn't think about it. Because of the whole map, you can also create a path from point A to point B. You go through this topic, that topic, and then you land on the topic you really wanted to cover. So yay, AI will help with it. So the science on one side creates a map, but the AI on the other side understands it. Fuck, I have to go.

Jaryd Krause:

We should have these chats more often.

Michal Suski:

No, no, no, I'm leaving.

Jaryd Krause:

Take some notes.

Michal Suski:

Send me the recordings. Send me the recordings.

Jaryd Krause:

I’ll send you the recording before it goes live.

Michal Suski:

Oh yeah. This is nice. Thank you.

Jaryd Krause:

No, thank you for coming on and having a chat. I'm a visionary like you, and I love thinking about how to make content better for the user, no matter where they are. And then looking at it through that lens of, like, not just going from Google to—oh, you're having your mind blown right now.

Michal Suski:

Yes, yes, the light bulb. My attention span is like, I'm over. I’m done.

Jaryd Krause:

You're done. You want to quit this call?

Michal Suski:

No, no, no, no, no. Maybe there is another opportunity. Just kidding. This is so good. I can't stop thinking. I can't stop thinking. But I will try. I will do my best.

Jaryd Krause:

I'm sorry for getting you out there.

Michal Suski:

No, no, no. Thank you for sparking this. Absolutely.

Jaryd Krause:

Yeah. So if you can look at it through what's actually happening on the ground, somebody that's like, “I want to learn how to start fishing or start surfing, and we look at it through their lens, and then you compare it to what's out there in the existence of what's already created. There might already be missing parts that have not yet been created on that journey that AI can find. I can send you this recording as well, so you have it if you want to pick it apart and all that sort of stuff. But while we're talking about AI, obviously…

Michal Suski

: I'm still here.

Jaryd Krause:

Yeah, you're still here. Sorry.

Michal Suski:

I am drifting away. Sorry. I promised.

Jaryd Krause:

Obviously, AI is not going away, and with Surfer SEO being so heavily involved in this space and developing for AI, you're obviously seeing that you're not worried about websites becoming destroyed, and then there's a lot of fear, at least in this space, of buying content sites. People are worried about, like, “Oh no, AI is going to take over.

And I don't know if my site's going to be able to exist anymore. Or the one that I buy, is it going to exist anymore? ” And then you've also got Google's search generative experience and how it's basically doing the same thing as a snippet.

What are your thoughts on how the rise of AI content is going to evolve? And I don't want to get into fear because too many people are speculating based on fear. My question is, how do you see it evolving in a way that's amazing for us? Because I'm down for it, right? I was a bit hesitant at the start, whereas, like, AI and all this sort of stuff, it's a bit of a shiny object syndrome, and it still can be in many cases.

At the same time, it's changing. And I love change because it levels up businesses and sites, and it's going to level up the internet. And I think we need it to disrupt this so we have no more crappy sites out there. So what's your take on how it's going to evolve?

Michal Suski:

It opens up a ton of space for new SEOs, really. Because it's all about the SEO industry, it's kind of disruptive, changing a ton and evolving over the years and sometimes even over the quarters. So the fresh blood will be able to cope with people who have ten years of experience in SEO. And this is absolutely fine because it increases competition. It increases the demand for quality stuff, and it will eventually be good for the reader of the content on the web, right?

Just because the SEOs won't be clicking and spamming some keywords and some links like they used to, job done. We're lambos. But it will be more like everyone has to adapt—not only the greenhorns, but really, the old SEOs will have to adapt as well to the new reality, which is great because it makes progress, like you said.

And from this SGE and stuff—well, the snippets—this is literally the same thing. You can't really answer all of the keywords and questions with a snippet. As a reader, you want to understand the value of the product. You want to read something more personal. And the AI-generated answer won't let you dive deeper into intermediate surfboards. Okay. If you want to understand what the surfboard height is for my height, that's perfectly fine. And you can get it from SGE. But if you want to understand the nuance between one model and the other, SGE won't help you.

Jaryd Krause:

They’re not going to.

Michal Suski:

Yeah. And when it comes to fear and stuff, well, Google has the fear. And for me, it's more to please the shareholders that we also have AI, right? But imagine you wait 12 seconds for the SGE to generate; who the hell has time for it? I would just scroll and take the website because it's much faster. 12 seconds? Imagine how penalized your website would be in the SERPs if you'd been loading for 12 seconds. Come on.

And not to mention the cost. The results are cached. There are data centers, and the index is there. And it just is. They just ask. And in terms of AI, they have to freaking read the content on the go and throw it away right after the question, which is a ridiculously unscalable solution for them.

Jaryd Krause:

I think you're spot on when you say Google should be fearful. Because where it's coming from is that I have people emailing me and worrying around, “Oh, this SGE thing. People aren't even going to come to my website anymore.” And I got frustrated, I must admit, on one of my emails, and I did some caps letters and stuff like that in one of my replies back. And it was to the tune of: Google earns $160 billion a year, right?

Now, if you owned, say, a beer company and the company makes $160 billion per year, would you stop paying the workers that brewed the beer? No. Because they're the ones that are creating it, right? So you're going to reward them.

People with blogs and content sites are getting rewarded by Google through ad revenue. And if they're not going to get rewarded, people aren't going to go to their website, and Google stops sending them, people are not going to create content or sites anymore. So Google cannot kill the golden goose here. That's where Google should be fearful, and we as site owners should not be.

Michal Suski:

If you eat the cookie, you have the cookie, right? Eat the cookie and have the cookie. There's no way. There's no way. I mean, they will produce this SGE and maybe roll it out to just a few people. They will probably make it sort of clunky so not everyone adapts, but good enough for the shareholders to accept that, okay, only Gen Z is using it and all the older people, including myself, won't. And like these old habits, they die hard, really.

So we still have plenty of time to sell to people in their 30s, 40s, and so on. I wouldn't fear it. Maybe it will show a little downside, maybe it will flatten the growth curve because SEO was always growing like this, and maybe it will flatten up, but it already flattened up. Kids are not on Google. They are on TikTok. They are somewhere else, on Instagram and stuff. I don't even know. Because, well, my kid is three and a half, and I'm too old for that kind of thing. So I don't know where they are, but they are not on Google searching for things, really.

Jaryd Krause:

Until they might get to an age where they’re like, “I just need a straight up answer for this, and then the SGE is going to give them a straight up answer. But it's not going to be as in-depth as they actually need. So they're going to click on the site, and the site owner is going to make ad revenue.

And it's the same with the cookies—the third-party cookies that they've just delayed and delayed and delayed rolling out. People are afraid of that. Why would they roll out the cookie unless they've got another better option that makes them the same amount of ad revenue or more? And they just haven't come up with that. So they haven't rolled it out yet. And they've probably been lucky that AI has come out and distracted people away from this a little bit. So they have had to put a lot of resources into AI now. But yeah, I think people's fear can be removed if they follow the money.

Michal Suski:

Yeah, exactly. I mean, there's nothing to be afraid of, at least right now. There can always be another atomic bomb by the end of the year, of course. But if it's not there, just don't be afraid of it because maybe it's not even coming. We shouldn't be worried about the future. We should always present and say, See, okay, there's AI, fine. I accept it. There's AI. So what can I do about it? I can speed up my writing process.

So I do. And I'm thriving right now. What's tomorrow? I don't know. I shouldn’t fear it. I just used the opportunity I have today. There's no point in killing today for future fear. Really, don't do it. Not only in SEO, but in general, don't do it.

Jaryd Krause:

Mic drop. We're going to have to pause the episode there and finish it up. That was too good. Spot on, Michael—Michal, sorry.

Michal Suski:

Oh, that's fine. I mean, both are. I'm used to it. I'm used to it, and it's absolutely okay. Don't feel sorry about it. It's Michael, Michal, Mixaw, or whatever, really. I'm really super happy to be here. So nice.

Jaryd Krause:

I'm so grateful that you have come. Guys, check out the link, Surfer SEO. There'll be a link in the show notes. Is there anywhere else that we should send people to check out some stuff about Surfer?

Michal Suski:

No, just your link. Your link, Jaryd, whatever.

Jaryd Krause:

Yeah. Thanks, man. That'll be in the show notes.

Everybody who is listening, thank you so much for listening. I really do appreciate it. I don't typically ask you guys to subscribe, review, and all that sort of stuff. If you want to, great. I'm not going to say no. But what I would say is to give the gift of giving and share this podcast episode. Yeah, it's great for Michal, Surfer, and myself and all that sort of stuff.

But most importantly, share this because we talk about AI and there are a lot of people with content sites or wanting to buy content sites who are worried about it. And us having this conversation around how valuable it's going to be and why we should lean into it the right way is going to be so valuable for them to hear this podcast. So please give the gift of sharing this. And yeah, we'll speak on the next one.

Michal Suski:

Thanks, man. Have a good one.