Greg Brooks - AI and the future of SEO

Greg Brooks: No. Hey, Tim, how's it going? I hope you're happy that you're having a wonderful day. I don't have anything to add. I think this is a really interesting topic. I think it's one of the examples of, of how technology really impacts what we do. And I think that we have this very predictable human response to it, which I'm sure we're going to talk a lot about. 

Tim Butara: A great intro and a perfect note to open the conversation on, but I actually want to maybe take a step back and because, you know, ChatGPT got publicly released in November, 2022. There's been a lot of development, a lot of discussion, a lot of conversation around that. And so let's start our discussion by determining what was the state of AI in search engine marketing and search engine optimization before the public rollout of GPT in November 2022? 

Greg Brooks: Yeah, that's a great question. I'm going to take it in a couple of different directions. So the first is in terms of what existed in AI technology, we were seeing basically hints of that happening. Like if you have Gmail, all of a sudden, one day, in Gmail or Google Docs, your sentence started getting suggestions for if you wanted to complete it.

If you were searching on Google, you know, about your sentence getting completed for you many different times before, all of those things are sort of AI driven. ChatGPT was really interesting because that was the moment that it became real for most consumers because it felt magical, which is kind of like in the hype cycle, that's the innovation trigger part of it. 

In terms of what was going on in SEO, in search engine marketing prior to ChatGPT, it really just depends on what your sort of awareness was. So at SearchTides, we were previously working with like DaVinci 3. 0, which is an LLM model. Basically, LLMs are essentially predictive programs that will determine what the next case of anything is. So what's the next number in a sequence? What is the next color in a bunch of colors? ChatGPT predicts, what is the next word that I should be saying, and then what's the next word after that that I should be saying? 

So they're all based on these large language models, LLMs, and they've... ChatGPT 3.5 was still using a lot of Da Vinci as its core LLM, which is publicly available, which search size was using in our own internal projects prior to then. So I think what happened was this public attention came on this thing. And for an entire industry, people said, okay, how do I use this? Or they said, oh, this is sort of the end of everything that I know to exist today.

And it went from there, but I think that's kind of what was happening before is all of these things were already in the background. They were in the tools that people use. They were in the apps that people use, and this is still search related and SEO related. It just became conscious for a large group of people at that time.

Tim Butara: So what about since then, what have been like the most important developments in terms of AI and SEO since November 2022? 

Greg Brooks: I think the most relevant one for search engines specifically is basically the onset of SGE. So, search generative experience. And that's when you type something into Google and it says, hey, we're going to kind of come up with this automated response for you that we think would be actually relevant to what you're looking for. 

And this was Google's way of kind of coming up with a predictive model. And everyone kind of said, everyone, everything is going to change forever now, because the search results are going to change indefinitely forever. All of a sudden, there's this new block of text that comes up. 

For us that was kind of just another example of how the search results have changed over the years. So Google has always tried to keep people on their site. ChatGPT always wants to keep people on their, you know, on the OpenAI website. Every business wants to keep the individuals who are using them in their domain.

And Google did this over time with search engines in different ways. One day, Google My Business showed up and you can make a call directly to a business with a phone number directly from Google and you never left. You could learn about a business directly with reviews and with photos. 

Then one day rich snippets showed up and that was what we called position zero. So all of a sudden you'd ask a question. There would be an answer at the top of the search results and you could step over everybody else in the first page of the search results. 

Then one day, this thing called People Also Ask popped up, which are questions that are related to the questions that you're asking, and all of a sudden, they're in the search results, and you can generate all this traffic.

So to us, SGE is just the latest iteration of what Google has been doing for at least like 10, 12 years. And it's just a... their more sophisticated AI driven attempt to do those sorts of things. People Also Ask, AI driven. Rich snippets, not as much AI driven. So there's been a little bit of an evolution in that process anyways, but I think the most important development is not actually SGE, even though it is dramatically changing the search results.

I actually think the most important development is the realization that you as an individual can utilize this technology to go way beyond your abilities previously as an individual. To me, that's the most important development. To somebody else who's looking externally, they might say, hey, this kind of search result changed forever.

We've been watching search results change forever. We don't even consider search engines to just be like Google and Bing. We think that ChatGPT is a search engine of sorts. You're asking any question, you're getting an answer to the question. You're using it to perform research. It's giving you an algorithmic response. They're all really similar. They just have different faces and logos and names and interfaces. So it feels different for people. 

Tim Butara: So you said that SGE, that's short for search generative engine, right? If I remember correctly. 

Greg Brooks: Experience. 

Tim Butara: Experience. Okay. Yeah. So... I thought that I probably had something wrong, but I'm wondering how specifically these search generative experiences can be used compare to like the traditional search engine experience and consequently, has the line between like a search engine and an SGE provided by something like ChatGPT become blurred, become kind of, kind of diminished? What's the situation here? 

Greg Brooks: That's a great question. And so SGE in this definition is specific to Google. That's the name of their product or their feature that creates an AI driven search result when you put something in. And in terms of what intuitively you would think, well, what's most likely to go into this answer box of this SGE, intuitively, you would say, well, something that was already at the top of the search results. Because Google has said if it's at the top of the search results, it probably makes sense for it to be at the top of this new response. 

But what we're actually seeing is that the overwhelming majority of SGE search queries are actually not answers that are found on page one. Which is really, really interesting. And I think that is freaking a lot of people out, but my response to that would be two things. 

Number one, you know, we're all familiar with beta projects. I would call this an alpha project. I would call this pre beta. So we don't necessarily know what things are going to look like in the future. But number two, this is the exact same thing that we just talked about in that list of new features.

Google My Business, all of a sudden reviews were probably the number one reason for you to get shown at the top of the search results, which was.. That changed overnight, rich snippets, all of a sudden, schema and the inclusion of tables and ordered lists allowed you to jump over the top 10 search results and get into position zero.

People also ask, all of a sudden building a page about something that was not the question that someone was looking at in Google, allowed you to show up for a thousand, two thousand, five thousand different search queries. That was not what your page was answering the question of. 

So there's been these disproportionate changes consistently for search engines. And this is just the latest version of that. And so the opportunity is who can kind of stay grounded and common that and say, well, what is the systematic approach? What are the sorts of answers that I need to be thinking about, as a website? And how can I look at this as an opportunity to really like put some jet fuel on the visibility that people have into the company, my project, my business, whatever it is.

Tim Butara: I'm guessing that this is why it was so important for like companies that are leading players in the search world to also develop their own kind of AI based search tools like Bard and Bing and whatnot, right? 

Greg Brooks: Yeah, it's a great question. And so I think like ultimately, all of these big tech companies have been waiting on the sidelines. So we're just going to keep using Google as an example, because it's really, it's just an easy theme to keep using. We talked before about stuff being in Gmail, about stuff being in Google docs, where it could finish your sentences. Like that is literally predictive AI. That is exactly the same mechanism that ChatGPT works and ingest the data set. It gives you what it believes to be the most likely response. And in this case, it's ingesting your data set and is giving you what it thinks this data set is most likely to say next. So that's literally the exact same thing as, as ChatGPT. 

Those things stood in the background for a company like Google, because there's basically regulatory concern. If we do something with AI. We have a Congress. We have Washington, DC, who's pretty litigative towards us, who's pretty heavy handed in terms of a regulatory component. And so we don't really see this incentive to innovate in this new area because we're already the market leader. 

Then think about something like OpenAI. They have no market share whatsoever. So there's nothing for them to lose. And also they haven't angered anybody and they haven't gotten sued by their government or by the European Union or anything like that. So they have more of a clear ability to build things out because their company doesn't have the same level of value. So what's there to lose? 

And so what we saw was you saw this with Apple, you see this with Microsoft, you see this with google, you see this with Facebook, all of these companies have really giant AI technology that just kind of was there. And it was only developed through one lens. Facebook telling you who it thinks your friends are most likely to be. That is an AI, that's an AI algorithm, but we don't think about it like that because we're not just asking a question, getting an answer from it. 

So all of these companies had these things, but because they had such large market caps and company valuation, they had to be pensive and they had to be more conservative with how they rolled out a new age of technology.

The big moment that happened was, okay, well now ChatGPT comes out and all of a sudden people are going there to ask it questions. And that kind of signaled, well, if there's no regulatory pressure against them, which there is, but not in a meaningful way right now, then that's kind of like a clear, that's a clear lane. Now these other companies can come in and do the same thing safely because they can look at the government and say, well, you've done, you've taken no actions against this company, so like you can't... the risk of you taking an action against me is way less now. 

Tim Butara: I love how all of these discussions around AI, it's always inevitable to mention at least at one point, how a lot of the things that we are calling AI right now, we're calling new now, and we're treating as new have actually been in place in our digital experiences, have been actually a major ,key part of our digital experiences long before November, 2022, when ChatGPT got publicly released. But I love... that's perfectly coincided with what you said at the beginning, right, about how that was like the point where it, it, it got from being something magical to something real. I love that. 

Greg Brooks: Yeah, no, absolutely. And it's the same thing with the people who interact with these things. If you were to, you know, previously it was, you were a data analyst and then all of a sudden we started calling people data scientists, and then we started calling people machine learning developers, and then we started calling people, AI engineers, and it's all the same stuff right now, 90, actually right now, all of AI technology is still prompt based.

So it is still a human being providing a set of instructions to a program running it and then basically course correcting indefinitely with that. So ChatGPT, what you don't see when you put in the question is you don't see the prompt that OpenAI has provided to ChatGPT prior to your question. So it's setting boundaries around do this, not this, answer,this, not this, look in this way with this database like this. And then here's the next question coming in. 

I have friends who own AI companies and they are, you know, building out different algorithms and things. They have basically... like the really, really good users have effectively a half a page of instructions for a prompt. And then they have about half a page of the negative instructions, the reverse instructions.

So we see these things as like doing things that humans are unable to do, but right now it is quite literally dependent on a human doing it and course correcting consistently over and over again. And that's why you get these really verbose prompts. AI, like prompt engineering, I think became a meme at some point this year because people were like, if you know how to prompt an AI.

But I think what people, what like we now know is that's actually... it, it is a meme. It's like, it's ridiculous that some dude's gonna come in with like a keyboard over there, you know, like a pull on keyboard and just be like, here we go, prompt away. Because that's... because ultimately the process is very manual. The process is, I put something into an interface, I get a response. It's most likely not what I'm looking for. And then I have to change my prompt and refine it over time. 

And that's how we've built the AI tools that we use at SearchTides. It's like, there's AI developers are really patient people who will sit there and iterate over and over and over and over and over again, until you get the consistency that you're looking for on a response.

Tim Butara: So, now I think that we're now pretty much caught up with like the state of AI and search, and I want to take a look at the future at this point in our discussion. I'm wondering what are your main predictions in terms of AI and SEO, maybe for this year and maybe for, let's say the next five years or so.

Greg Brooks: In terms of SEO, we are basically entering an era where at SearchTides, we say the cost to produce good content is going to zero. So right now, you kind of have all of these articles online about people being like, here's how I program ChatGPT to make an article and it's okay, but it's definitely way better than it used to be. And it is definitely a few iterations away from being pretty darn good. And it's a few iterations after that, away from it being pretty darn good for many people to actually do quite easily. 

And when you hit that point, how valuable is good content in terms of it being informative in terms of it being educational? The answer is not that valuable because it's actually really easy to produce.

So what then becomes valuable and at SearchTides, what we say is what will become valuable, we call it, if we look at the past, we call that the foundational layer of SEO. Those are all the things that everybody is really familiar with. That's stuff like link building, that's stuff like having onsite optimization, that's stuff like having good content, all the pages, that sort of thing. 

The layer that we're in right now is what we would call the middle ground. And this is the influence era of SEO. And that for us is, yes, your authority is important. Yes, your trustworthiness is important, but your expertise is important and your experience is important.

So Tim, I'm actually going to give you an example of that. We use it as pretty good. Here's what expertise and experience is. I go into a pharmacy. And I say, hey, I have this prescription for this thing. Can you fill it? And the pharmacist says, yes, I can. However, there's actually this generic version that's like 10% of the cost. 

The pharmacist is using their expertise to know that that is something that also exists, even though a doctor prescribed you X, you will be way better off with Y because it's the same thing. You just pay way less. 

Alternatively, I can go to the pharmacist and I can say, hey, I have this prescription for X, and they say, you know what?

I totally get it. That's what everybody walks in here with. Everybody who I give this to, it doesn't work for them. They come back, and then I give them this other thing, and then that works for them. Why don't you just get this thing now? It'll get you a better result. That's the pharmacist using their experience to be able to create a better result.

Those are the human elements. If you think about the human experience, it's what's the culmination of all the things that I have encountered as a human being? And how do I talk about that? And how do I showcase that? And that is different than education. And it's different than information, which is what an LLM is fantastic at.

We shouldn't be competing in that area. 

Well, how can I do better than you going to an interface, asking a question and gives you everything that you want in two seconds and spits it out all there for you? Like, why would I try to be a better version of that? 

What I should do is embrace the human element, which is why the future of SEO, we're calling it the human era.

And there's a lot of technical things that are going to be happening in it, but in terms of the broad strokes, us embracing our humanity, our individualism, what makes us unique as a species beyond just being an encyclopedia of information, that's the old way of thinking about it. Who we are, what we've encountered, what we've learned, and what we bring forward, those things are going to govern what people are looking for when they're just not looking for encyclopedic information. 

Tim Butara: I love that answer. So I'm guessing that people working in SEO shouldn't be super worried about losing their jobs because the future of SEO will be like driven by humanity.

Greg Brooks: Yeah, it's a great question. So you can always be worried about losing your job. I'm going to talk about why that would be the case and why that's not the case. So number one is we have for sure seen that SGE is changing what's in the search results. I'll give you an example. So another partner at SearchTides, he loves Halloween. He makes this huge Halloween house. Everybody in his community comes and hangs out. And so he wanted to make a house of mirrors this year. 

And he said, he went on to Google and he searched for how can I make a cheap house of mirrors. And what he got back were instructions about how to cut wood and he's like, well, that doesn't make sense. And intuitively, it doesn't make sense. But then the string of connection is actually, if you want to make a cheap fun house, you have to use glass and wood. You can basically get glass online and you could order wood online. But you're going to have to be able to cut the wood. If you're going to do it cheaply, you can't pay somebody else to do that. So you need to know how to cut wood. 

And so that's like, I think that's an example of an overreach and something that wouldn't exactly be that in the future. I think if you ask Google that in the future, that's not the response you're going to get. But it is an insight into how this thought process exists and that there is a thought process, it is simply just different than what was there before. 

And so if you can meet that level of prediction, the way that we think about it at SearchTides and what we recommend to people, is we say, Google loves Wikipedia, they love Wikipedia because number one, it's structurally very sound from a technical perspective. It is fully cited. So you don't need to trust anything that they say, that you can verify all the information that's being said. They do a great job with stuff like interlinking, but also they hit up all the topics and the subtopics within a given area. 

Your website should do the same. We have this idea that like top of funnel, middle of funnel, bottom of funnel, let's just focus on conversions. The reality is you need.... if you are making your website, your company the hub for all things that had to do with your industry or your sub industry I'm not saying like, oh, now I have to go be The Encyclopedia Britannica of finance.... like, no, inside your sub industry inside all the things your customer needs to know across all their parts of the customer journey and being able to predict the needs of your customer or your visitor the same way that SGE can say you need to know how to cut wood if you want to make a cheap fun house. If you can do that hetn you're really future facing especially because you're overlaying it with your humanity. 

Now if you're just trying to present information and skate around and your website is kind of like A lot of super generic top of funnel content, yeah, you're in trouble because that's the stuff that an LLM can absolutely do better than a human can, because there's nothing unique about that. That's just gathering information and that's not really the area that we think it makes sense to compete in. 

So there is risk from that perspective.

But it typically comes from things are already going well for you. And therefore you have the most to lose as opposed to you have a lot to potentially gain. If you're not at the top of the search results for everything, you should be thrilled that this has happened because this is your chance to get into the top of the search results.

Tim Butara: Nice. 

Greg Brooks: Yeah. 

Tim Butara: Man, I really love this discussion, Greg. Just before we wrap it up, I'm wondering, based on everything that we've discussed so far, as well as like on your work with clients, what you've seen in the industry, what you've seen with like maybe your peers who are also working in the same field, what would be your top tip or top few tips for marketers that are trying to win with SEO in this AI driven landscape?

Greg Brooks: Amazing. I'm going to list one to eight different tips. I don't know how far we're going to get before all the ideas are flying off. So we were talking before about the human future of SEO and what that means. And so tip, you know, one tip is i, we call it writing from the heart at SearchTides, literally speak from your human existence, your experience, your expertise, what you bring to the table.

Number two, use technology to give you comprehensiveness. So what else do I need to talk about on this topic? You can put stuff into ChatGPT. You can put your top three competitors into ChatGPT. You can say, what are the different topics and subtopics that this block of text is talking about? You can put your page into ChatGPT and say, do I have all of those? Are there things that I need to add on here? And so you can use technology to give you completeness. 

The next tip is remove fluff. So we have a metric at SearchTides that we call value per word. I think previously with content on the internet, there was a lot of, how do I make this 2000 words instead of 1000 words? There's a top seven list. How do I make a top 15 list? And that's not how people absorb information. It's actually the opposite. 

So how can you trim down the fluff from your article and how can you speak more efficiently? Is there something that takes you three sentences to say, but you can actually state in a half a sentence? Do that throughout your entire piece of content that you're writing.

And the next big tip is link building is going to matter more and more in the future because content is going to be easier and easier to create good content for, which means that everyone's going to be able to do it. So what's the thing that's going to matter more? Third party endorsements and that's links pointed to your website. And so therefore link building is going to be more and more. 

That depends on basically what type of business you are. If you're a local business, you want to generate local links. If I'm a local car dealership, I want to go to my local chamber of commerce. I want to go to my regional car company, you know, whichever brands I sell, and I want to get a link from their websites. I want to support local businesses and sponsor things and have them link back to my site, because you're going for relevance in a specific geographical area. If you're a local business. 

If you're a national business, you're basically playing an authority game, which means you need as many high quality links as possible at scale. Obviously we do a lot of that at SearchTides, but the tips are basically to get links pointed to the pages that you care about and the pages that link to the pages that you care about as well. Pretty impossible to go through like all the tips and tricks for a link building. We have a lot more content about it at searchtides.com you can check it out, but ultimately those kind of frameworks. 

Number one, you know, embrace the human component. Number two, enable technology to help you and to guide you and to assist you and to further what you can do. And number three, like understand what will be easier and then therefore what will be more valuable in the future. That's how we've been working since, you know, January, 2023 from this methodology. And it's worked really well for us. 

Tim Butara: These were some fantastic tips, Greg. I'm sure that our listeners will get a ton of value from them. I know that I certainly will. Just before we wrap up the call, if our listeners would like to reach out to you, get in touch with you, learn more about you, where can they do all that?

Greg Brooks: Thank you for the opportunity to shamelessly So SearchTides is a hub for all things SEO. If you are tangentially related to SEO, if you're in marketing, if you get website traffic, if you are in SEO, if you lead a department in SEO, if you're on product, anything that touches SEO, you can come to searchtides.com and you can learn all about how we think about these sorts of things for free. 

We're also building a lot of content on social media. So check us out on LinkedIn, we SearchTides on LinkedIn, check us out on YouTube, we're SearchTides on YouTube, we hold monthly digital events. You can learn about them more at searchtides.com. You can hop on our email newsletter. That's probably the best way to make sure that you're staying on top of everything and that you're just getting kind of this passive.Information out to you about all the things that we're doing. 

And we do also have an agency arm. So if it makes sense to, if you want to bring on an agency and you think it might make sense to have a conversation about that, you can get in touch with us through our website as well.

Tim Butara: Awesome. We'll make sure to provide all the relevant links in the show notes for easy access to you and Greg, thanks again for a great discussion. This has been fantastic. 

Greg Brooks: Thanks so much for having me, Tim. 

Tim Butara: And well, to our listeners, that's all for this episode. Have a great day, everyone, and stay safe.

Outro:
Thanks for tuning in. If you'd like to check out our other episodes, you can find all of them at agiledrop.com/podcast, as well as on all the most popular podcasting platforms. Make sure to subscribe so you don't miss any new episodes, and don't forget to share the podcast with your friends and colleagues.