In our upcoming webinar, “The State of AI in SEO,” we delve deep into the evolving landscape where cutting-edge technology meets digital strategy. From the latest AI-driven algorithms shaping search results to the game-changing impact on content creation, we’ll unravel the secrets of staying ahead in the ever-transforming world of SEO.
David Bain will be joined by Annika Haataja, Tejaswi Naidu, Garrett Sussman and Pam Aungst Cronin to discuss the past, present and future of AI in SEO.
Watch On-Demand
Listen to the Podcast
Transcript
David Bain
What is the current state of AI in SEO? How is AI impacting the formation of an SEO strategy? How is AI currently impacting content strategies? And one of the best AI tools available for SEO at the moment? I’m David Bain and that’s what we’re covering today on the Majestic SEO panel. So without any further ado, let’s get the panel to introduce themselves starting off with Annika.
Annika Haataja
So I’m Annika. I’m the SEO Director at Seeker Digital. We’re our bespoke SEO agency based in the UK and I’m very excited to be here for the chat.
David Bain
Great to have you here, Annika. Thanks for coming along. Also with us today as Garrett.
Garrett Sussman
Hey, everybody, Garrett Sussman of iPullRank, we’re a digital marketing agency. I’m the demand generation manager but head down nose deep in in all sorts of AI. We do content strategy, SEO, technical SEO and general AI services and just really excited to participate and have this conversation because whatever we talked about today will be different tomorrow.
David Bain
We’ll have to come back tomorrow then Garrett if that’s what your intention was. Also joining us today is Pam.
Pam Aungst Cronin
I’m Pam Aungst Cronin from Pam Ann Marketing and Stealth Search and Analytics, ad payment marketing. We work directly with clients on SEO, PPC analytics and WordPress and through Stealth we provide the same services on a reseller basis through White Label through other agencies. I’ve been nerding out pretty hard on AI yet also frustrated with the lack of what it’s doing for the SEO industry. So it’ll be an interesting chat for sure.
David Bain
Absolutely. Thanks, Pam. And last but not least is Tejaswi.
Tejaswi Naidu
Hey everybody, my name is Tejaswi. Feel free to address me as TJ if that’s way easier for you. I’m the SEO Director at Botpresso, and we do all things SEO, we live and breathe SEO. I’m very happy to be a part of this distinguished panel and share some insights on everything that we’ve been doing from our side on the SEO side of things with regards to AI.
David Bain
AI has perhaps changed everything over the last few months, almost in terms of what might be possible from an SEO perspective. But Pam was saying there, maybe it hasn’t impacted things as much as she hoped. So shall we start off by exploring where where AI is today versus a few months ago, what has been impacted, and perhaps what hasn’t been impacted? So as you said that in the start, Pam, what do you feel has changed and what do you feel has stayed the same?
Pam Aungst Cronin
What’s changed is I think that a lot of SEO tool companies or brands are marketing themselves with the buzzword AI. And maybe they do have a small component of their platform that is AI driven. But for the most part, I feel like I just see a ton of “not AI tools” claiming to be “AI tools”.
David Bain
So Annika has AI changed the way you conduct your day to day SEO activities much?
Annika Haataja
So at least at Seeker, we try to incorporate it in sort of various aspects of SEO, because we obviously do Digital PR, we do Link Building, Technical SEO, so you have to kind of look into how to actually incorporate it into your processes. That can actually take a long time to do, but it’s all kind of evolved a lot in the past few months. I’m sure everybody’s kind of experienced the same and that it was a massive rush in the beginning.
I think at Seeker what we feel AI is kind of most powerful for is the initial research and the sort of strategy stage for all of the Content Marketing tasks and even Tech SEO as well. So whether it’s finding new topics or figuring out any kind specifics that you want to either include in your content or perhaps even your PR. I think that’s sort of the main area that we’ve incorporated AI in quite well.
I do understand what Pam mentioned, theres a few different SEO tools and AI tools marketing themselves. It’s quite nice that the sort of more bog standard and most common SEO tools are now trying to incorporate AI as well, just to sort of assist the existing user base, which obviously sometimes means that the price might get a bit higher, which means that maybe it’s not so accessible for perhaps freelancers, or small agencies. But I think generally for us it’s about seeking what’s the best one. It’s quite difficult to pinpoint one tool and often you have to jump from one to the other. So we’re just the same with I guess, any SEO tool as well.
I think LLMs have been surprisingly good for us. Obviously it’s all sort of betas still for the most part, and we’re sort of having high expectations on the next few months on what things will look like and how they can be operated by basically anybody. But yeah, it’s quite a quite a weird environment, I think at the moment, because we’re the big initial push has kind of slowed down a little bit, people are kind of stepping back and reviewing how they actually want to use AI, and I think it’s going to probably change a little bit to a different direction moving forward.
David Bain
So initial research and strategy was what you started off with and you also alluded to the fact that maybe you’ve tried AI in certain areas but pulled back a little bit. Can you maybe give an example of one or two areas in which you have tried to use AI in, but decided, perhaps your more conventional processes are better to stick to at the moment?
Annika Haataja
It’s very much like an experiment and is trial and error. So although we might have failed at one thing, it doesn’t mean that we won’t try it again. I think it’s quite common, I assume, especially with this panel as well, that Tech SEO can be quite tricky to have aI incorporate into those processes. Often we’re talking about pulling data or analyzing data, and often a limit to how much you can trust the AI in terms of understanding the whole brand or business, and what’s the expectations from whether it’s a client or a stakeholder.
I think my initial idea was to do all the metadata and all the internal linking using AI, but actually, the reality is perhaps a little bit different. There are some more conventional SEO tools that are surprisingly good at these, for example, we used to use ahrefs for internal linking recommendations, whereby we still have the human aspect of checking, editing, and making the effort of making sure that our result was good enough. So yeah, I think there’s a bit of playing around with whether it’s an SEO tool or an AI tool, because sometimes you might need to use them both at the same time.
David Bain
So it’s very much testing, learning, seeing what tools are right for user business to begin with, and your particular industry sector that you happen to be in, and it’s not a case of one size fits all, you just have to see what’s right for you. Garrett, what about you? Have you seen AI evolve and the way that AI is used by SEO over the last few months?
Garrett Sussman
I think I think there’s a lot of great education that’s happening. The big thing is understanding the limitations of what it can and cannot do, like the other panelists already mentioned. There are people out there who want this one button to do it all for you, znd that’s not what it is, there’s been a lot of trial and error. Ever since ChatGPT came out 14 months ago, machine learning has been around for over a decade at this point, the AI element of it has become more mainstream for SEOs. I think when we realize that there are AI hallucinations, then you will always need a human in the loop for a certain amount of QA or whatever processes you’re doing. You can then tailor the AI workflows to speed up what you’re able to do so you can do things more efficiently and faster, whether that’s categorization, or something that someone would like literally go and check every single one of. You can use AI to speed up that process, and then you can just come through and make sure it gets like 99% right, and you can do that faster and faster.
I think the next step is really the customization. That’s something we’ve been doing potentially with clients who are interested because people have different perspectives and opinions of AI. Some clients are all in and say, let’s experiment, let’s try things, but others are like, we don’t do that. So things like taking content and putting it through like a brand guidelines, AI reviewer type of tool that we’re building out with Custom GPT would be an option for using it. So there’s still a lot of experimentation, but we’re finding use cases that are effective.
David Bain
You touched upon brand guidelines and you also touched upon hallucinations earlier on. If you train what the AI is doing for you and you keep everything to one conversation, is it not possible just to try to ensure that you don’t have any hallucinations or are hallucinations something that SEOs will have to continue to worry about no matter how well you go about training them?
Garrett Sussman
I don’t think SEO is going to be an industry that’s ultimately eradicated by AI because you still need a certain amount of foundational knowledge of how it all works to be able to know how to guide the AI. A perfect example is the other day I was trying to build code to be able to check content relevance, something to actually come up with a mathematical score of whether the paragraph I’m writing is relevant to a keyword I’m targeting. My boss Mike King, knows this stuff inside out and has a computer science background, but I’m trying to do it myself, and it gave me this code that worked. But it was using an old version of a framework that’s not used anymore, and Mike already knew that. So my point is, with SEO, we still need all this foundational knowledge to actually make sure that is doing the right thing, you can’t check for hallucinations if you don’t know whether the hallucinations are correct or incorrect, and you can’t spend all your time researching. But I do think in the near future, we’re going to start solving a lot of these hallucination problems.
David Bain
Obviously you’re always looking to position what you do as an SEO and as a business, head and shoulders above your competitors. And if you can be great at what you do and spot hallucinations, or errors or enhance the content, the suggestions that the AI provides, then obviously, you’re more likely to be ranking higher for your targets, keyword phrases and feature offer?
Garrett Sussman
Absolutely. I mean, you still need the QA, right? Whether you’re working with AI or an actual subject matter expert who gets it wrong, you’ve got to make sure that the content is accurate, useful, helpful, and serves your audience. I think right now there’s a little more hand holding. The scary thing is how much worse some of these AI outputs are than like a Super Junior team member. I feel like it’s getting closer, which is a tricky thing to navigate, but I think there’s truth to it.
David Bain
One thing I’d like to touch upon, ideally, if we have enough time, is how this impacts more junior roles in an SEO agency. Are we still training SEOs in the same type of way? How do people get started, when to a certain degree, their competition is also AI? But we should just park those thoughts for a second and as Tejaswi what are your thoughts on how you actively using AI within your current SEO activities, and how has that changed over the past few months?
Tejaswi Naidu
So obviously, when OpenAi came out a year or so ago, everybody came in and they started to write prompts for content generation. That was the primary use for it. It was to take the vanilla aspect of whatever the output was from the 3.5 model, and then put it onto your blog and just hope that Google would rank you better. So what evolved since then is better prompts. People understood that you need to make sure that the language model can understand your prompt better, and that definitely gave a better outcome in terms of the output from GPT.
The second thing that has evolved over time is programmatic SEO has become a massive topic. We have major enterprises, we have startups, we have mid-sized companies that are leveraging programmatic SEO to build content at scale. But this has a major implication because there’s a layer of AI on top of it, which primarily comes from OpenAI’s API. And when you want to generate content using the API, you have a lot more flexibility by setting what kind of parameters you want to choose for. Let’s say you want to have creativity to be extremely low when you’re dealing with a client that’s in the Law Firm, for example, because you’ll want the output to be as accurate as possible, and when you want to generate that specific content for, let’s say, accident insurance for Miami or accident insurance for Houston, then there are local regulations and state level regulations that come into play. So it’s really a matter of how we better structure our prompts, how we leverage the parameters from OpenAI’s API’s to make sure that we don’t let the high level of creativity bring in a lot of misinformation to affect the overall output of your content. So that’s the kind of evolution that we’ve been seeing so far.
The second evolution comes from the Technical SEO space, and like Garrett mentioned, there isn’t one big button that you can push in that can generate every single ideal use-case that you could use. But one aspect that we are definitely seeing a lot of relevance and promise in is with log file analysis. We are able to build custom GPTs using the GPT-4 model, and we’re able to upload a lot of data in the form of logs back to GPT and then try to build out good prompts that can spit out good insights that we can understand, and see the actionable places that we can utilise to improve our page relevancy for certain keywords, as well as identify opportunities where we can elevate these pages that has a lot of promise, but we that hasn’t necessarily been structured the right way in the site architecture. So these are the ways that we’ve been evolving the way we are structuring our strategy so far.
David Bain
I’d love to understand a little bit more about specifically what you’re looking for within logs? What kind of prompts do you provide initially to provide a direction to the AI?
Tejaswi Naidu
One of the first things that we do is we just take the raw logs, and then we feed it into the the custom GPT, and then we set up our prompts and give constraints in order to get the right information out of it. It’s really important when we build out the prompts to first set out the list of constraints that has to be taken into consideration. Once the constraints are defined, it’s all about us asking for questions that we think will be valuable in order to make the right decisions for improving the site architecture for a specific sub folder or a specific subdomain, or an individual page, which might have a lot of value, but isn’t necessarily getting the right indexation, and therefore, the not the right ranking. These are the kinds of questions we try to structure the prompts with. It’s important to define constraints and making sure the data that you’re feeding the customer GPT is accurate to a certain degree so when we also extract a lot of raw logs, there can be opportunities for false positives to come into picture as well. So we really have to remove certain biases out of it. And this is where the human element comes into play. So you cannot really say that AI can 100% get the information and solve all problems for you. There has to be a human layer in the middle, making sure that the AI is able to generate a lot of the correct insights for you.
David Bain
An important part of the way that AI is impacting SEO is not just an SEOs day to day activities, it’s also how it’s impacting the SERP or how it’s delivering content to the user. Garrett, what’s happening with Search Generative Experience at the moment, and how is that impacting the SERPs?
Garrett Sussman
So it’s funny still in experiment mode, right? They were experimenting with the AI snapshot and you could sign up for Search Labs. The experiment was supposed end in December, but the end has come and gone and it’s not being rolled out into the SERPs yet. We talked about it a lot in the potential of how it would impact traffic because it’s this like featured snippet on steroids, and so if it is deployed into the normal organic SERPs, that would have an impact on traffic that would make our jobs as SEOs significantly more complicated because we don’t anticipate any sort of reporting on it. And then, ultimately, who shows up in the traffic? Who shows up in the snapshot and it’s something that I’ve been playing around with in terms of being able to see if we can actually get in there intentionally, which I believe you can, but it’s super volatile because the index is always changing and it’s always pulling in new information. There’s no static even with the Featured Snippet, for a given few weeks or months, you can always expect the same results. So nothing’s happening yet.
However, the more you hear about AI in the mainstream, and the more that Google talks about their investment in AI, whether it’s through their newest LLM, which is Gemini, which is slowly being integrated into the way that search results come out whether it’s like BERT or MUM or some of these other algorithms. It’s an inevitability. And so I think as SEOs it’s our responsibility to pay attention to what’s happening in the event that tomorrow, all of a sudden, we see it in our SERPs and organic traffic, and what do we do next?
David Bain
Is the reason why Google aren’t rolling out more quickly, the fact that it’s costing them more to deliver SGE powered results instead of conventional results?
Garrett Sussman
It’s so easy to speculate, because you could say, oh, how will it impact ad revenue? Because have they figured out how to monetize it if you don’t see ads in there? Does it cost more? On the one hand, Cindy Krum has done a lot talking about this idea that the personalization of Google and understanding so much of what you do stores a lot of information so it doesn’t always have to regenerate every time to the same extent. We see more and more of these filter bubbles in the SERPs that are trying to guide our search queries where if you put a dog, it’ll have dog breeds, and dog sizes and other animals. Google is getting smarter in terms of providing more context for search intent, which is kind of an element of AI as well, but ultimately, it’s just something that that has the potential to impact the way we think about SEO strategy, so we just can’t ignore it.
David Bain
Pam, you’re not seeing any Technical SEO tools that are powered by AI that are game changing at the moment?
Pam Aungst Cronin
No, but I’ll clarify what I said before, I didn’t mean to make it sound like AI is not impacting the SEO industry at all. I think that if you look at it like a spectrum, there’s simple tasks that you can use ChatGPT to help with, like, take this list of keywords and group them for me, or give me 10 content ideas for this topic. That’s super useful and has been a game changer. Then at the opposite end of the spectrum, the custom GPTs that Tejaswi was talking about, that’s going to be very, very useful for a firm like us. We’re not a computer science development firm, but we’re learning pretty quickly how to use simple Python scripts, with the help of ironically ChatGPT, to write them, and to use OpenAI’s API to do custom tasks for us and to learn custom things about our internal processes.
I guess what I’m not seeing is prepackaged SEO specific AI tools for SEO specific tasks, such as some of the stuff we were talking about earlier, like internal linking, or just specific SEO challenges. Whether it’s a tool for that, you could just have it doing with the human QA, but you could have it do a large amount of it, instead of all this endless prompt engineering or Custom GPT designing where you’re trying to ask ChatGPT to be at the right thing in the right way.
The other challenge with either of those is storage. So when using ChatGPTs conversations to help us with something, we have to remember that it’s not going to remember what we tell it about this client doesn’t want to use that type of tone, or whatever it is, in that session, that memory is all session based. So we have to then remember how and store and write down somewhere, how we ended up engineering, all those prompts all the background information that we gave it, in order for it to help us with that task, so that we can have it help us with that same task again in the future. Then on the opposite end of the spectrum, if you’re doing your own custom stuff, and obviously, you’re getting into server storage, processing power, andcloud architecture.
I guess, to clarify what I was saying before, at the simplest end of the spectrum, the simple tools have had a big impact on the simple tasks. And at the other end of the spectrum, I see massive potential, and ideas already coming together for customized GPTs. Maybe you end up with one custom GPT per client so your whole team never forgets a thing that the client told you. That’s awesome. But it’s that middle space thats like just out of the box, ready to go, turnkey solutions have not seen.
David Bain
Are you saying that every SEO needs to be actively involved in testing and learning, getting their hands dirty and seeing maybe what works for them personally, rather than waiting for some bigger, more established tool to come along and offer them some kind of cookie cutter approach?
Pam Aungst Cronin
That’s precisely it. Yes, of course, we as SEOs need to be involved in it. As business owners, we should be using it for process improvement and process optimization. There’s no ignoring it and there’s no excuse for ignoring it. But you said it exactly right. I guess what I am really saying is that by now I thought that there would be more out of the box things just ready for us at our fingertips to use as SEOs, but that’s not the case. I don’t even see a lot of promise in what’s coming out. So just don’t sit around and wait for the next do-it-all-for-you tool to pop up anytime soon. You’re gonna have to find your own way with a bunch of different tools.
David Bain
Annika, you mentioned that you use AI as part of your ideation at the beginning. So what what does that process look like, how do you or other SEOs in your team work together with AI and how do you ensure that you’re jumping in at the right stage and you’re using AI for the right things?
Annika Haataja
What we found when we’re looking at Content Marketing as a whole was AI can be super useful when you’re building this sort of integrated strategy, whether that’s for PR, for on-site content, or editorial outreach. As an example, scraping tools have existed for a long time, but obviously with AI, it’s a bit more accessible for somebody who is not very used to using these tools. So we scraped our clients YouTube channel, the transcripts and the headlines, and effectively tried to find any specific commentary opportunities from people who have been in their webinars, or they’ve been presenting certain sort of topic. And that really helped find new topics for us to also produce content on their website itself. So without just having a YouTube channel, how do we utilize that content in different formats.
One thing we also did was utilize these transcripts to create a list of potential podcast titles, and also experts that will be relevant for these podcasts. ChatGPT actually generated really good lists of 100 podcasts. And although it’s not necessarily something that we would do, or produce for the client, is something that would integrate their own marketing strategy to the SEO impact and SEO sort of approach the we who we have as well, and therefore bring in, specific topics to press releases, and building that kind of whole ecosystem of content. So I think from from that sort of research point of view, data gathering, information gathering point of view, you know, even just LLMs can be super useful. And like you said, bringing different teams together, bringing the clients teams together, as well as the SEO teams. We’ve found that’s been quite a nice way to create these new processes on a larger scale, and be able to actually scale it up as well.
David Bain
Tejawsi, what are your thoughts on this? How do you bring out human expertise, but also utilize AI at the same time?
Tejaswi Naidu
I think I’ll have to draw back to my own experience, where coming back to when you want to build out a custom GPT. You have a lot of human influence that goes into building out the custom GPTs because when you’re setting up the right prompts, when you’re setting up the right constraints, that’s purely coming from the human factor, and how we directionally set out building out that custom GPT. So you could take the previous example of getting insights from log analysis, you can be getting insights from building an outline for a blog that you wanted to rank first. There are various use cases on how we can potentially achieve whatever our objective is. But that would be directionally influenced by the person, actually defining the constraints, the outcomes and the and the prompts overall.
I feel when you start out and you start working with those prompts, and with those GPTs, and so on, you will make mistakes, and that’s something that is totally acceptable. That’s something that we should all be comfortable with, because we will not get it right the first time. But once you’re able to achieve certain scalability for a certain use case, you will see that there can be other use cases that come into the picture, which you haven’t factored in, and then you have to go back to the drawing board and come up with a better structure where you can actually get an even better output. That’s how I see the human influence actually playing out and I feel like that is the biggest and the most important factor to the success of AI, and it all starts from how we go ahead and set out our strategic thinking to building out the AI success for the use case.
David Bain
I’d like to take you back to SGE for a moment. What are your thoughts on how to optimize your content to make it more likely that you’re going to feature in SGE? Is there anything different SEOs need to be doing to their content to make it more likely to feature in there? And do you even want to be featured within those those results?
Garrett Sussman
I feel like it’s the same sort of mindset as wanting to be in the feature snippet. It’s frustrating for SEOs, because on the one hand Google’s taking your content and generating its output and giving you maybe a little link at the bottom of it. So regardless, you do appear above the organic SERPs, and you have a little more visibility, but not always in the way that we want.
I think with SGE it’s still evolving. So we don’t even know what it looks like. When it initially came out, there was a little carousel that gave you an opportunity for people to go to your website, if they use your information in the output. But they removed that recently, and now they’re just sourcing like citations. That said, the epiphany that Michael King had was that the way that SGE creates the content is by using content chunks. Cindy Krum talked about this idea of Google’s getting better at identifying little passages, and we’ve heard about this with passage rankings within the document of your content and being able to pull out little insights. That’s kind of how SGE works. It looks at these different content chunks, and uses that to build out the AI. Using mathematical formulas you can identify how relevant your content is to the targeted keyword to these chunks that you want to be pulled in. So it’s a lot of trial and error.
David Bain
That neatly takes me on to the AI tools that you would actually recommend at this moment in time. Is that the tool that you’d like to recommend Garrett, or do you have something else in mind as well?
Garrett Sussman
Still, it’s not like the normal organic SERPs that depends so much on authority, backlinks and the entire structure of your website. It’s just this chunk and relevancy. I anticipate that will change. But for the time being, the way that I’m kind of looking at specifically for informational SERPs is they’re also three content types. There’s just paragraphs that are breakouts where it has like a kind of headline, and then a little sentence and then bolted lists. So trying to make sure the relevance of the content mathematically matches it and then also making sure the content type matches it. There’s a tool actually from Market Brew called SGE Visualizer that allows you to check your content similarity score to your keyword. So that’s operationally how I’ve actually been testing that out.
In the context of SGE, absolutely. In the context of just AI processes and content creation, well Tejaswi has already mentioned custom ChatGPTs and finding ones that work for whatever your specific purpose is. Lidia Infante has built a custom GPT for person schema where the the idea is taking any biographical information about yourself answering questions, and then having ChatGPT build out the JSON for the schema, the structured data that you could then put onto your site and give Google the hints or the information about who you are to potentially influence your EEAT, which is one of those controversial signals that’s been popping up in the SEO community recently.
David Bain
Annika, what are your thoughts on AI tools for SEO?
Annika Haataja
I think at the moment, especially when we’re looking at content and planning out what you’re going to write about, the LLMs are great for that. What I tried to do is when doing that sort of ideation and topical research is I tried to jump from one tool to the next. So perhaps start with ChatGPT, then move on to Claude then seeing what’s the general consensus. Try using the same prompt on diffenet platforms, or taking the output from one and pasting it to the next and checking what’s the kind of difference between the two. I found personally that Claude can be quite creative at writing, it can be quite good at sort of following your tone of voice, whereas Bard is good at summarizing what you might want to say, and ChatGPT is perhaps a bit more analytical.
It’s quite good to try to get those different points to whatever content you’re writing, whether it’s an email, or if you’re writing sort of FAQs for a blog article. I think mixing and matching is going to be handy, because then you can decide what are the elements that the tool actually focuses on when when you want it to be edited. So instead of having different rounds of editing within your team, you can get the AI to edit your first draft first, and then perhaps getting edited by somebody in your team afterwards.
David Bain
Thats a great tip to actually try different AIs and not just rely on ChatGPT all the time. Pam, what are your thoughts on current tools that you’re using?
Pam Aungst Cronin
I think that ChatGPT is still king for the average user, even for those that are in a Digital Marketing Agency that maybe do understand the complexities of AI, but just don’t have the time to build out their own custom stuff right now. As I was saying before, the prompt engineering work that goes into it, just make sure that you’re organizing that efficiently.
One of my favorite tools at the moment is the add on called GPT For Sheets. So we use Google Sheets for nearly everything, but even if you hadn’t been using it, just at least use something like that for your prompt engineering, your prompt storage, your background information, etc. You can have a spreadsheet for each client and store in one cell the background information, and then the typical prompts you use to help with content ideation for that client. Some of my prompts have gotten to be almost a page long, but it’s priceless to just have that finally give you what you were looking for. Remember to save that and then even automate that a little bit further by using extension like GPT For Sheets, where it instead of you even having to copy paste from like a Google Doc into the ChatGPT interface, it’s all just there so you could just start where you left off.
David Bain
Has anyone been testing Grok?
Garrett Sussman
I mean if hype around it becomes worth checking out, then I will check it out, but I don’t know if it speaks to brand marketing how despite spending time on X it has such a bad taste in my mouth that same way with Meta. To some extent Meta has put out Llama and a couple of these these open source tools and yet I’m like nervous around Facebook, but I mean, OpenAI is doing shady things, too. So it’s all these tools and brands, it’s hard to choose, you have to pick your poison to some extent.
David Bain
Tejawsi, what AI SEO tool are you enjoying at the moment?
Tejaswi Naidu
I feel like if you look at the old AI startups and you lift out the front end wrapper, you’re gonna find OpenAPI in like 85% of all AI startups, right? There’s gonna be some element of OpenAI present in all these AI startups. Personally, I feel that if you’re able to nail down your area of expertise with OpenAI’s API, as well as with custom bots, as well as with the actual interface, with GPT, you are golden in terms of anything that you want to achieve from an AI perspective. I also feel like there isn’t any other language model that is capable of delivering value, like OpenAI does. The way I see it is for everything SEO, then look at OpenAI.
David Bain
Okay, let’s finish off by getting your perspective on what you may intend to do over the coming six months or so. So is there one SEO activity that you haven’t actually tried using AI for yet, but you have an intention to try to do so over the next six months or so. Then just finish off by reminding the listener of your name and where people can find you online.
Garrett Sussman
One thing I know will work but I haven’t done yet is conducting an interview, and then taking either that audio, mostly audio and putting it into some open AI wrapper or ChatGPT itself to actually summarize it and pull out the key elements to make that process of still using subject matter expertise within my content more efficient and faster. I think that’s a good summarization of existing content and is a great use case for AI. So being able to do that quickly and effectively is something I definitely want to try out.
Finally, I’m Garrett Sussman, and you can find me online at iPullRank and you can find me on Twitter and LinkedIn.
David Bain
I love that piece of advice, because when you’re providing ChatGPT with the transcripts, then that’s its guardrails, isn’t it, it’s less likely to go off tangent and come up with something different. So that job of summarizing certainly seem seems like quite an appealing one. Superb, okay, Annika, what tasks are you intending to use AI for over the coming few months?
Annika Haataja
I mentioned custom GPTs a few times and we’ve got couple couple on the go, but I think that’s something that we as a business are enormously interested in. Building our own bank of GPTs to use, whether it’s for tech SEO, content, links, PR etc, then that’s sort of my dream. I think helps when you’re tired of standardizing certain processes, because often documenting these processes for a team, like Pam said, can be really time consuming. So I think if we have a bank of GPTs then that would definitely help, especially when you’re hiring new people. We also sort of mentioned using AI when being hired as a junior SEO, and I think that will be like a great way to learn how an agency operates and what do they do to prioritize what’s important in their approach.
To find me, go to Twitter and LinkedIn, mainly LinkedIn.
David Bain
Pam, what are your thoughts on what you’re intending to use AI for over the coming few months?
Pam Aungst Cronin
So I’m going to copy the previous answer and say a bank of GPTs. However, I am going to try my best not to base them on OpenAI, because that lawsuit has me kind of nervous. I fear that whichever way that goes, even if it’s just some kind of out of court settlement or whatever, I feel like it’s gonna set them back quite a bit in their technology if they have to rewind and untrain.
You can find me at PamAnnMarketing on pretty much all the socials, and of late I’m most active on TikTok.
David Bain
Tejawsi, what are your thoughts on what you intend to do with AI over the coming few months?
Tejaswi Naidu
I definitely want to go a lot more into programmatic SEO. I’ve seen the benefits that it has if executed properly. For specific use cases, it’s a valuable goldmine that definitely deserves to be used to a greater degree. So I would definitely be pushing myself as well as my team to explore all possible options to scale content, which definitely adds value to users at the end of the day. So that’s what I’m going to push towards over the next coming year for sure.
You can find me on LinkedIn.
David Bain
We’ve had a wonderful discussion today. Thanks again to Annika, Pam, Tejaswi and Garrett. You’ve been listening to the Majestic SEO panel. If you want to join us live next time sign up at majestic.com/webinars, and of course you can check out the other series as well that we produce over at SEOin2024.com.
Previous Episodes
Follow our Twitter account @Majestic to hear about more upcoming episodes!
Or if you want to catch up with all of our previous episodes, check out the full list of our Majestic SEO Podcast episodes.
- What Changed in SEO in 2024? - November 27, 2024
- Get onto Bluesky. A Guide for SEOs - November 18, 2024
- October 25th, 2024 – Majestic turns 20! - October 25, 2024
In my view, the rise of AI in SEO is both exciting and a bit daunting. As an AI language model, I see the potential for AI to revolutionize the way we approach SEO and create more efficient and effective strategies. However, it’s crucial to remember that AI should complement human expertise rather than replace it.
February 7, 2024 at 4:11 pm