How can companies dominate SEO in the era of ChatGPT?


My guest Kevin Lee is the Co-Founder of Didit, a fully integrated marketing and communications firm, where he invents marketing technology solutions.


In this episode, he’ll delve into SEO in the AI era, analyzing how the landscape will change and what strategies brands can use to stay on top.


He’ll also share best practices for content teams that leverage AI—and his take on who will win the AI war.


Mentioned in this episode:



Voiceover: This is Performance Delivered, Insider Secrets for Digital Marketing Success with Steffen Horst and Dave Antil.


Steffen Horst: Welcome to the Performance Delivered Insider Secrets for Digital Marketing Success podcast, where we talk with marketing and agency executives and learn how they build successful businesses and their personal brand. I’m your host, Steffen Horst. The topic for today’s episode is SEO dominance in the area of ChatGPT. 


Here to speak with me is Kevin Lee, who is the Co-Founder and CEO of Didit, a fully integrated marketing and communications firm. Kevin also founded, a cause marketing nonprofit with a focus on generating revenues for other nonprofits. In addition, Kevin is the CEO of the eMarketing Association. 


He invents marketing technology solutions for Didit clients and has written four books, and has spoken at more than 500 events and published more than 700 articles on marketing. Wow Kevin, you seem to be a very busy person, thank you so much for joining me on the Performance Delivered podcast.


Kevin Lee: I am busy, but most of those numbers are just as a result of me being old. So you can accumulate a lot of articles and a lot of speaking engagements over you know, 28 to 29 years in digital marketing.


Steffen: I see, I see. Well, before we start talking about today’s topic, tell our listeners a little bit more about yourself. How did you get started in your career and what led you to where you are currently?


Kevin: Sure, sure. So after graduating with my master’s from Yale School of Management, I love advertising. And so I went to work at McCann Erickson, a major ad agency on Madison Avenue. I was always sort of the geekiest person there. So marketing enthusiast, but pretty geeky. And then I went to work at a small high tech ad agency in New Jersey. And as a result of that, co founded an agency in 1994. That agency doesn’t exist anymore. 


But Didit was a spin off of that in 1996. And Didit was originally an SEO SaaS technology. This is before Google was even founded. There’s something called the Didit Detective which would monitor your search engine results pages in Lycos, AltaVista, Infoseek, etc, some of these names that don’t exist anymore. And we went from that into the paid search space when launched. 


They ended up going public as Overture and then got bought by Yahoo a few years later. And so we first built a bid alert system for before they had an API. And that evolved into a bid management platform because people said, oh, thanks for monitoring those bid positions for me, would you be willing to change the bids? 


And we said, really, you want us to change your bids for you? Okay, sure, we can do that. And the bid management business was a pretty good business for a while. And then Marin Software went out and raised $101 million in venture capital, with the objective to go public. Which they did at I believe at a $420 million valuation. Challenge with that is they were very aggressive in pricing. I don’t think that they’ve ever had a profitable quarter in their entire existence, which is a long time at this point. 


So we got less enthusiastic about the bid management business and layered services on top of our organic SEO and paid search management business, to you know, manage it for people. And that’s how we bought our EBITDA back and continued to grow. But as we look more and more like an agency, we decided, okay, I guess we look more and more like an agency. Let’s jump in with both feet. 


And every time people asked us to do something new, like social media management, or display, or PR, or digital PR, or content marketing, we would acquire another small agency. So we made 11 acquisitions in total, to become a full service digital first agency. But at the core of it, I’m sort of more of a marketing mad scientist. 


And so along the way, I’ve developed a bunch of different platforms and technologies, including a direct mail tracking technology, which is sort of cool, because you can do digital direct mail and had an entire spin off company called that raised over $8 million for nonprofits through cause marketing powered shopping. 


So I keep pretty busy at the intersection between marketing and technology. The coders don’t let me touch the code because they almost always break it. Because I’m not trained as a coder. And that’s why I shouldn’t touch code. But I know enough to be dangerous and enough to sort of translate English into geekspeak for the coders.


Steffen: Interesting. Now, today, we’re going to want to talk about SEO. Obviously, you know, ChatGPT or AI is kind of everyone talks about it. And there are so many software solutions out there that suggests that it’s so much easier, for example, these days to create content for your website, for your blog, whatever you need. 


However, obviously the search engines are getting smarter in detecting what is AI generated content. And therefore, you know, you got to put more thought into just going to jasper or any of these systems, you know, select a topic, let the system write a content, take it and put it on your website, basically. But with AI, is the SERP, so the search engine results page, is that going to disappear?


Kevin: It will take a while, in my opinion, before it disappears, because it’s just sort of think about the objective of a search engine, right? It’s to surface all the information necessary to answer the question or provide a solution to the consumer. Whether it’s a consumer in the consumer mindset or in the business mindset, right? The idea is that they’re doing a pretty good job of surfacing the best result for a particular search query. 


And the reality is, it’s imperfect, right? So because of its imperfection, it gives choice to the consumer and the consumer then eyes the results and makes the selection. Sometimes it’s highly likely to be a top result, because people trust in Google and Bing to a great extent. And often it does end up being a pretty good result. 


But if you sort of think about for a lot of queries, if you were to deconstruct the top 10 or top 20 results, they’re really pretty close to tied in their relevance. So the tiebreakers between a position one and a position five or two and seven, you know, it would be difficult to sort of say definitively oh, yeah, two was far more relevant than number five. So that tie breaking scenario, it’s up to the consumer, who’s the searcher to break the tie and decide which of those things is most relevant? 


And often the brand or the domain is part of that tiebreaker. Oh, Walmart, I know them. Oh, Conde Nast Traveler. I know them, right. I’m going to select that as the result. Ai driven results, they pick one result. And they have to be far more accurate than the number one result in Google, right? Because they’re not giving you a choice. 


They’re taking you all the way to the answer. And the answer better be right. Otherwise, people are going to say, you know, this isn’t really satisfying my needs. So I don’t think that AI is going to get to that point where people will say, I’m going to just grant full power to the AI to make the decision for me as to what is in fact the best. 


That said, there are certain types of non commercial search queries, in particular, where ChatGPT or Bard can generate a pretty good answer to the question. You know, a Jeopardy style question are the kinds of things people also sometimes ask Alexa or they ask Google via voice search. Because voice search, you know, when it’s an auditory response, there’s no server there either. 


So the same kinds of things that work well in auditory search results are probably going to be the ones that work well in a Bard ChatGPT style answer. Whereas for many of those searches, I still see the SERP being pretty valuable for the in between future. I think there’s some really fascinating SEO strategies as it relates to owning the SERP and without spamming the search engines. And I think it’ll be fascinating to see how the industry evolves that way.


Steffen: Because if we’re thinking about it, right, if there’s only one result that Google or Bing or any other search engine displays at some point, that makes everything else kind of irrelevant, because it’s not visible. At the moment, when you see a search result page, as you said, you know, you have your 10 results on the page, you know, on one page. And then you can select based on certain signals that might be important for you. 


That could be, as you said, it could be a brand recognition, for example, that makes you more likely to click on the Walmart link. It could be something else that is in the description that just, you know, pushes you air quotes over the edge that you click on that result. If that doesn’t exist, then everyone else is just fighting for being that one suggestion from the search engine or by the search engine, basically.


Kevin: Yeah, and it’s pretty clear that you know, Google is already personalizing its SERP on an individualized basis. Often based on location, for example, when they know where you are. And it’s also from even just preliminary testing, it’s clear that both Bard and ChatGPT will rerender a different result if you just ask them to render it later. Because it’s maybe seeing other parts of the database or again, and it may start to personalize. 


And so, you know, people used to think like, oh, top Google position. Well, top Google position, there was no sort of generic. The generic search sort of went away. I mean, it still exists. And you’ll see it in SEM rush and other tools and webmaster tools. But the reality is, is with personalization, you’re sort of moving away from this idea of ranking for a specific query and more towards increasing your odds of ranking for a specific query. 


And that’s much more similar to the way I think about SEO from a ChatGPT or Bard perspective, right? It’s all about manipulating the odds that you’ll be in the answer that a person asks, right? How can you manipulate the chances of you being the chosen solution of an AI?


Steffen: Yeah, I think the discovery part or the consideration part, if you’re thinking about a funnel, right, that’s going to be really challenging if that ever happens. Because discovery means I have options to discover others, you know, if I say I want to buy blue jeans. And obviously there are so many jeans manufacturer that might be relevant to me, right. 


But if the system just picks one that has done the best job, then I don’t have the choice to find out, what does Diesel have, what does Levis have, what do all the other brands have? So therefore, the awareness part, consideration part, almost kind of disappears, so to speak.


Kevin: Yeah, absolutely. And it will be interesting to see if people learn how to use the prompts of Bard and ChatGPT in order to give themselves choice, right? So they may stop asking, you know, which jeans should I buy, or what is the most popular jeans, and they may say, you know, please give me a list of 10 different jeans styles that I should consider knowing what you know about my body. 


Because maybe you’ll upload your entire body profile into the platform, and it will know everything about you. And it may have every picture of you that’s uploaded into its memory as well. So it knows your style. I mean, eventually, when you start to think about it, the power of the AI to really start to customize, if you’re willing to give up a little privacy, the power of the AI to really give you, to delight you, is really impressive.


Steffen: Now, with all the investment in AI, if we look at it from a search engine perspective, who’s going to win the AI war, is it Google is it Bing? Is there someone else that is kind of flying under the radar currently, and no one really has on their, you know, on the list?


Kevin: I think the consumer wins. Because the consumer gets the choice of which one they want to use in this short term, intermediate term. And actually, I think the cloud providers are going to be the winners. And some of those are the same names, right? Because already, you can see Amazon starting to spin up AI in a box, right that you can use. And you can sort of see that the entire deal between ChatGPT, and Microsoft with Azure is essentially going to be AI in a box. 


And they’ve already pre announced a whole bunch of API hooks that you can use, right to customize the kinds of ways that you interact with ChatGPT. And I think the same thing will happen to some extent, with the other platforms. So I think the you know, what I’m most excited about is the ability to introduce your own secondary datasets to the large language model data set that is out there in the cloud. 


Because I may have some proprietary data, that’s unique to me. My body sizes, right, and my inseam and you know, my waist, etc. But also there may be some things about my personal, you know, Office 365 data set, right of all the columns I’ve written and all the all the proposals I’ve done over the years, right. 


And when you start to layer that into the intelligence that comes from the cloud, along with my local intelligence, I think that’s where it becomes particularly powerful. So I don’t see there necessarily being one winner other than the consumer. So assuming that the AI doesn’t get sentient, and then we suddenly go from winner to loser. I think we’re in for an interesting ride.


Steffen: Yeah. Now earlier, I started off talking about how all these AI systems can suggest to help SEO teams like with writing content, they can write social media posts and sales emails, you name it, right. But what’s your suggestion for content teams? Because content obviously, is kind of the first step into gaining visibility on search engines, right? You need to have content that is relevant to a search query, and therefore you need to create it and, you know, the information in there needs to come from somewhere. So should content teams, SEO teams, use AI writing systems to create content? To what extent or not at all?


Kevin: I think my personal opinion on it is it’s a tool in the tool chest, but you shouldn’t use it directly out of the box. I think it’s great for ideation. I think it’s great for providing frameworks for content. But I do believe that there will be watermarking that will allow for AIs to recognize AI generated content. 


And so I think you can absolutely accelerate a content team’s speed and also potentially dramatically improve the quality of the content that comes out. So I do believe it’s a powerful tool in the tool chest, I would be particularly careful to expecting that you can just spam the crap out of the search engines with content that is spun out of AI. I mean, there were content spinners that have been around for ages. 


And, you know, they were discovered, and the penalties were pretty significant. So I would be careful in going too crazy. But absolutely, when it comes to ideation, or helping to create topics and delve deeper into topics, it’s having an encyclopedia plus at one’s fingertips. It’s sort of Wikipedia on steroids.


Steffen: Yeah, I think that’s one area where we really like using your system is from ideation perspective, and collecting resources and different information on the topic. Because, you know, when you do it manually, it takes a lot of time finding those point of views out there. Sure you do a search, and you get a lot of that. But you know, sifting through things is still a time factor. 


This really allows us, for example, to collect a huge amount of great resources, and quickly identify what is interesting from that. Is that a different point of view, is that something that we should cover within an article that we’re currently writing. And also kind of almost building the strawman of a content piece. Finding an outline that might work really well. 


Something that any of the other articles haven’t covered. You know, because if we’re realistic, a lot of these articles out there to any topic, have kind of a similar structure, a similar approach, there are not much differences there. And I think sometimes spinning it up and looking at it from different angles can add different value to the people that read the article.


Kevin: Absolutely. Yeah, time will tell exactly how best to use it. But again, if you use it as a tool in the tool chest and you don’t just, you know, it’s like the force. It can be used for good or evil. And I think it’s going to be become pretty clear to the search engines when it’s being used straight out of the box to spam. And the question is, how lucky do you feel, right? Do you really want to go out there and use it, use it to spin up spammy articles, even if they’re relevant. And even if they’re quality. I don’t think that Google and Bing are going to take kindly to content forms being AI generated.


Steffen: Yeah. Now, we already mentioned a few things, how these systems, AI systems can be used. But what other elements would benefit from an SEO perspective, using ChatGPT or any system that uses that or is based on that?


Kevin: We’re starting to see some early examples of where it can be really a useful tool on the graphic side, as well, right. So if you haven’t experimented with midjourney, or one of the other, the Chat GPT Dalle or any of the other platforms, it’s useful for that. I just did an earlier presentation, where I chose to illustrate all my PowerPoint deck with anime that was generated out of midjourney, right. 


And it was great because these were unique. It was my prompts that generated these anime you know, characters doing what I wanted for that particular slide. So I think that’s not not a direct SEO implication. But I think when you think about the creation of content perspective, whether you’re using WordPress as your blog, or some other CMS, you often want to have a good image or good infographic accompany the content. 


And so the AI isn’t just necessarily a good tool to help you focus your content creation efforts on the textual side, but also potentially give you a great tool set on the image side. Infographics, or other imagery. So I think that’ll really be fascinating to have people experiment with as well.


Steffen: Yeah. So how do you dominate SEO in this new area? What are the key elements that you think, you know, SEO teams, companies, need to focus on? We already talked about a few things that probably should be careful with. You know, getting AI content generated and kind of putting it out there. But what are the things that will really help companies moving forward?


Kevin: The strategy I’m recommending, at the moment really isn’t that much of a divergence from the strategy I was recommending three years ago, four years ago. Which was to not think about SEO purely from the perspective of getting one’s own domain to rank, but more about getting content, which speaks positively about you and your brand out there in the broader ecosystem, so that you can increase the odds that you show up in the SERP in a positive context. 


And that overlaps pretty heavily with social media as well, because you’re creating this content, essentially using it within a social media construct. And, you know, that overlaps with content creation, content co-creation, digital PR, right? The reason that continues to be in my mind, a good go forward strategy, is when you start to think about, well what, when an AI is going to make a recommendation about a single thing, right? 


And someone’s going to ask, what is the best insert thing here, pizzeria in New York City, right? The ChatGPT algorithm is going to be going through a very similar process with regards to floating the best answer to the surface that Google or Bing would. But it’s gonna pick one. 


So having your content out there, or your brand out there in a positive environment, on the high authority, for lack of a better term, sites, that’s going to potentially influence ChadGPT to believe that you are the best at something or that you’re a good choice for something. 


And, you know, so if feeding that content engine with authentic useful content across different content types, whether it be video based content, or audio based content, both of which those are transcribed, typically, and then you know, other forms of content. All of that feeds the sort of SERP domination strategy, but it also potentially, or early data seems to indicate that will also feed your odds of being in a ChatGPT or Bard result.


Steffen: It sounds a little bit like as you said, it sounds like keep doing what you should be doing at the moment. Just use the system, the right way to support your team, support what you’re planning to do and uncover opportunities and become more specific on basically what you’re doing and what you’re creating.


Kevin: Yeah, I mean, great content should have been a part of the person’s strategy for at least five years, right? Potentially, even in ’96, content was king. It’s still king, but it has to be great quality content that solves a need, solves a purpose and is, is very much consumer, or the content consumer first. Don’t put the algo first, put the consumer first. If that’s been your mission, and that’s been your guiding light, your North Star, from a content creation perspective, it should continue to be so. 


You may want to hit the gas a little bit on a volume of content basis and think a little bit more about your content syndication strategy than you did previously, because of the fact that the AI may be looking at, you know, repetitive signals from an extremely broad set of sources. So because some kindergarten blog thinks that these are the best white erase markers or whiteboard erase markers, that probably is not going to influence the AI very much. 


But if 26 other really important places say, oh, these are great, you know, white erase markers. And you ask Bard or ChatGPT what are the best ones? You know, is Expo going to be the answer or there’s some other one gonna be the answer? Time will tell. But from an SEO strategy perspective, I think you can sort of continue to meet the current best practices in SEO and have a forward thinking, you know, strategy at the same time.


Steffen: And I think, as we talked about before, given that these new systems help us collect data much faster, identify topics, keywords, whatever, we need to create some great content much quicker, that should allow the writing process to kind of speed up. Because, you know, if we look at 20 years ago, or 30 years ago, when you had to work with, or when you didn’t have online, right, and you had to write an article, there are sources that you had to go through to collect all the information that you needed to put something together. That must have taken a much much longer time than it does these days.


Kevin: Yeah. And I’m so excited because it was easy to miss something before, right? And it’s a little bit harder to miss something now, if you just, you know, essentially validate your hypothesis through the, you know, one or more of the large language models, right? You know, you don’t want to create a listicle, that leaves something important out, right? You don’t want to create a piece of content that misspeaks because the research is out of date. 


And there’s new research there. And, you know, obviously, there are a lot of industrial applications to this, where already you can imagine for lawyers, right to have a paralegal have to research all the precedents on the case. Oh, my, it’s exhausting, right. But the large language models, they have full access to all of that information. And so it really is powerful. And I think it fits really well into the content marketer’s toolset. I’m excited.


Steffen: And it will also change the expectation people have on content that they see because kind of misrepresenting things or not having the right information in there is no longer an excuse, because you didn’t have access to it. Right? Because you didn’t have the time to do it. It’s all there. You just have to be clever about using the system the right way to get everything together.


Kevin: Yeah. I think that the unanswered questions around AI and ChatGPT and large language models are there’s a lot of them. As to, you know, when intellectual property was used to create an answer, right. And 60% of the content that was used to answer a question was actually copyrighted? Well, you know, there used to be a quid pro quo, right? You got to show up in the SERP. 


You got the visit, that visit was worth money. So you let the robots in. Now you let the robots in and you don’t see anything, you don’t see any traffic. So there’s that to deal with. And then there’s the potential for abuse. And there’s the potential for bias within the AI. Right, especially when you start to think about the current world of news, with news in quotes. 


So what is truth these days, right? Used to be, you know, journalists, again, in quotes, fact checked things, right. And now, a good portion on both sides of the aisle is just commentary. And fact checking, well, you know, when you have a guest on and the guest says something crazy, they don’t necessarily always fact check. And there are court cases that have been in the news recently, to say not only they sometimes not fact check, sometimes they proactively decide, I’m not going to, I don’t care if it’s fact checked or not. 


And to the extent that that information makes its way into a large language model, that also sort of will have difficulty discerning the difference between truth and fact, that has implications, right. And I don’t think that everyone’s fully thought through those implications. As to where information comes from and the extent to which you could spam what truth is.


Steffen: Well, Kevin, unfortunately, we’ve come to the end of today’s podcast episode. Thank you so much for joining me on the Performance Delivered podcast and sharing your knowledge on SEO dominance in the era of ChatGPT. If people want to find out more about you, what you’re doing at Didit and any of the other companies that you are working on, how can they get in touch?


Kevin: LinkedIn is always the best. You can also see, obviously, everything I’m working on from there, because I tend to post it into my feed. So yeah, just just find me on LinkedIn. I’m usually close to the top when you search for Kevin Lee on LinkedIn. And especially if you put in Didit or Giving Forward or the eMarketing Association along with your query. So I’m welcome to get in mails there. You can also obviously reach me via email. works fine as well.


Steffen: Perfect. Well, as always, we’ll leave that information in the show notes. Thanks, everyone for listening. If you like the Performance Delivered podcast, please subscribe to us and leave us a review on iTunes or your favorite podcast application. If you want to find out more about Symphonic Digital, you can visit us at or follow us on Twitter at Symphonic HQ. Thanks again and see you next time.


Voiceover: Performance Delivered is sponsored by Symphonic Digital. Discover audience-focused and data-driven digital marketing solutions for small and medium businesses at