In memoriam: A tribute to Bruce Clay 

My heart sank when I learned that Bruce Clay had passed away. I’d heard he was in the hospital, but all I could think about was the two long conversations we had last fall — one to catch up, the other for a podcast interview.

I first contacted Bruce almost 25 years ago, when I emailed him cold to ask whether I could republish some of his industry writings on ethics. He said yes. My cited article unintentionally ranked #2 on Google for “Bruce Clay” for years. I’d often joke about that with him later, and he seemed both amused and annoyed—presumably because I’d done it with his own content and blessing.

A few years later, I worked with Bruce and many other search industry pros on the board of the Search Engine Marketing Professionals Organization (SEMPO), a business nonprofit focused on promoting the then-new search industry. Beyond promoting best practices and the legitimate business case for search, we later worked on U.S. Internet policy in the early 2010s.

SEMPO had board members from around the globe, and it literally took some of us around the world. I got to know Bruce through that work, and we’d later run into each other while speaking at conferences, sometimes even landing on the same panels. We had a great time doing great work. The organization lasted about 15 years, and if I remember correctly, Bruce was a founding member around 2000 or 2001.

I have a specific memory of Bruce from when SEMPO board members were heading back to our hotel on the east side of Midtown Manhattan after dinner. A snowstorm had just begun and would leave several feet by the next day. The usual loud traffic noise was dampened by the weather and the absence of most vehicles. It was eerie, but almost joyously quiet. The city that never sleeps was effectively taking a nap under a blanket of snow.

Then something happened that I’d never seen before or since.

As snow silently dumped into the streets, a massive lightning strike hit just a few blocks away, over Bruce’s shoulder. I’m not sure he even saw it directly. It felt like an explosion, and we stood there for a few minutes trying to process the contrast: a shattering bolt hitting the street between massive skyscrapers amid a torrent of snowflakes. But not a drop of rain.

None of us had ever seen or heard of it, and we didn’t know what to call it. I believe Bruce called it “thunder snow,” and it stuck. Thus his naming streak continued.

Bruce was, and is, the real deal in search. His legacy isn’t just that he coined a term. He consistently backed it up by pushing the field forward, educating others, and engaging with the people he loved. Like many early pros in the field, he helped develop search design practices that remain foundational today. He was known as a top expert through his writings, interviews, books, and hundreds of industry events. For many who remember the beginning, and even follow him today, he is the GOAT.

Bruce took an intellectual view of search, and I don’t think he always saw it as a job. It was exciting and new, and how often do you get to invent an entirely new thing? Not often. He recognized that AI is such a moment now and approached it with the same excitement and insight. Many of you in this industry may only now be learning that he pioneered things you do every day. They seem like common sense now, but they weren’t. There was a lot of debate just to establish the basics.

He wasn’t just passionate about search. He was also passionate and generous toward the people in search. If you were into it, you were part of his tribe. That was certainly true for the many thousands of people in the industry he interacted with, myself included.

We could always get deep in the weeds on the trade while also taking a wide, visionary view of where things were and are still going. He was an engineer with an MBA, and it showed in his leadership, expertise, and authority on the subject. He was that rare talent who understood the trade from top to bottom, and back to the top again.

And he was genuinely a nice guy, with many friends around the world. In our last conversations, I sensed he was content with his life and accomplishments, and felt blessed by the journey life had taken him on. He had nothing to prove.

In his podcast interview, he was as insightful as ever and delivered some of the most sensible thoughts I’ve encountered on where search is going in the new world of LLMs. He was as innovative now as he was when everything first started 30 years ago. Since much of search is about linguistics, I’ve been especially interested in how we think about, and what we call, this “new” thing. I agree with what Bruce had to say in the discussion below. The conversation crystallized my research, and the industry seems to have largely followed suit over the last year.

If you’re one of the many thousands of people who talked shop with Bruce over the years, read on. You’ll likely relive some of your own moments with him.

As I went through the podcast transcript, I realized we had hours of additional conversation recorded beyond search, talking about cars and other things. At the end of our first conversation, he said goodbye with great love and care. That’s just who he was. Those words speak to me differently now, and they always will.

Rest in peace, Bruce. You are already greatly missed.

Rob Garner
Talk about how you got started back in the 90s, and then talk a little bit about the naming part of of search. 

Bruce Clay
Okay, I’m going to drag us kicking and screaming back to 1996. I decided I wanted to switch corporate jobs to be a consultant looking for something to do. My background is math programming. I did mainframe stuff in Silicon Valley. I did PC stuff. I did networking. I did a lot of optimization work in that environment. That was actually my primary role was optimizing mainframes and PCs and applications and program products. In January ‘96, Al Gore really kind of pushed the Internet. And I listened to him and I said, ‘oh, look at that. It has marketing. I have an mba. It has technical work’. That’s sort of like my programming. And it just showed up and nobody was really able to figure out how to make money on it. 

I decided I was going to get in and start doing it. And of course in order to do it, you had to play with the search engines. So I started targeting search engines. What are they looking for, how do I update web pages? And automagically, they started ranking. Then people started calling me and they were giving me money. Here I am thinking I’m going to be a one person consultant. And it just took off. I mean it went vertical. And I now have offices in Japan, I have them in Australia, I have them in throughout Asia, throughout Europe, India. I’ve been to the Taj Mahal twice. I mean, you know, those kinds of things. And I would never have bet that it would have gone that ballistic. 

Rob Garner
It’s incredible. Back in those days there were “tens” of people doing it literally early on. You were doing sort of the agency side of things, doing this for companies and businesses – the “legitimate side.” There were obviously a lot of other folks that were doing black hat techniques. But you were really one of the first to set up an actual service for businesses to help them rank for in many some cases their own brand name, and in other cases generic terms. 

Bruce Clay
Yeah. Now this was also three years before Google. So you know, there it was a wild west when it comes to search engines. Everybody thought they could make a fortune and they were growing overnight. There were 20-some major search engines. 

It turns out they were all taking data from each other. The very first SEO conference, all of the leaders in SEO – and none of us knew how big this was going to get – but all of the leaders in SEO showed up for the conference and were sitting at one round table in the bar. You know, had there been a natural disaster, we would have wiped out the industry. 

Rob Garner
Which conference was that SES in Dallas? 

Bruce Clay
I don’t remember where it was. I just remember I’m looking at all these people whose names I have known because we all post and communicate and were all sharing, Danny Sullivan was the one that pulled it together for Search Engine Strategies. So that was pretty much the very first. The biggest contribution I made – I was very vocal about it being the future of the Internet from the perspective of businesses. I’ve been given credit for coining the term “SEO,” and you know, the fact is nobody knows who was the first to ever say anything, but I do know there were only a handful of us. 

And so we decided it was SEO and that was really the first acronym. It did stick. We all used it, we propagated it in our writings and everything else and it stuck and back a little bit. 

Rob Garner
Back in the day when Danny was writing for Search Engine Watch, it was really the first journalistic news site in this industry, which later became Search Engine Land when he sold Search Engine Watch. And you were also referring to the Search Engine Strategies conference, which was the first major national business conference on the topic. He’s obviously working for Google now, so he hasn’t been out there the way he was really for the first 20 years of search, since he’s been the Google liaison. 

The folks that are working in the industry now in the last five to 10 years, or maybe even 12 years, they might not have seen what a force he was in using that term. And it wasn’t a slam dunk necessarily, was it? I mean there were other terms that were being considered. Like iProspect and “search engine positioning”. And how do you think it came over to “SEO” versus “optimization” versus “positioning”? 

Bruce Clay
Well, there was “positioning” and there was also “ranking”. And so between the three of them, I personally think that “optimization” probably won because it was more of a technical. I’m able to get paid a lot of money for doing it. 

You know, it’s like fine tuning a race engine. Your daily car or a racer. And I think that optimization had more flair for people to say ‘hey, I want to do this for a living’. They hooked their wagon to that term and it just, as I said it went vertical – and I mean really vertical. I mean this was not a slow growth term truly like within a three month period. That is exactly what it was worldwide. I mean it was amazing. 

Rob Garner
It’s part of the core digital marketing. It’s obviously a core part of any type of earned media marketing tactic as well. So that “optimization” part has definitely cemented itself in the lexicon for sure in all of marketing. 

So the one thing I’ve been writing about lately is these new terms that are proposed or you know, AIO, GEO, AEO, but they all have the word “optimization”. The novel concept was coming up with the single word “optimization”. 

How do you feel about the application of that word to what’s perceived as “new” things with AI and maybe just talk a little bit about what you think about those terms and how you think optimization fits into that, this new area or new era we’re moving into. 

Bruce Clay
You and I have discussed this a number of times. I mean, if you recall, I made a statement that search engine optimization is in no way tied to organic. It doesn’t say “I’m only organic”, it just happened to be at the beginning. That’s all there was. But it doesn’t say “I’m only doing organic results”. 

We believe it’s anything the search engine spits out that causes business that can generate traffic. That is the optimization target doesn’t have to be organic. 

Rob Garner
Anywhere there’s a query and answer, and in a way it can be affected. 

Bruce Clay
Yeah. And that’s it quite frankly, what’s called AIO or AI as it shows up in search, it’s search engine optimization because it is output from a search engine that causes traffic. Now, my specific take on it is if you look at what is or what’s considered SEO, part of that is backlinks or linking. Links are a big part of SEO. There wasn’t an [acronym] to accommodate “link engine optimization”. I mean, there was never a need to invent new terms. I believe that the majority of the terms that are being created are just inventions to. It’s a marketing term. It has nothing to do with the practicality of doing optimization. They just invented a marketing term and said, I do this now, nobody else does this, you better hire me. 

And I think that every one of them is nothing more than an extension of SEO. Just like links are or on page or what have you. Google has even explicitly stated that SEO is the foundation of any AI results. That if you are not found organically, you know, if you’re not in the top 10, give it up. Right now there’s going to be a disturbance in the force, and maybe we’ll get to that a little bit later. But even with AI mode, you’re going to find that if you don’t rank organically, you don’t rank in AI. 

Rob Garner
Understood. And you know, it seems like anybody that uses AI, they see the citations coming up. It’s almost looking like AI is becoming for search and navigation and frankly, [it is only a] sliver of the real overall purpose and benefit of AI. It’s one segment of it. But we can see those citations right before our eyes. And it does feel a lot like SEO – the basics, and to what you were saying, I’m hearing a lot of what sounds like repackaged SEO tactics for this “new thing”. You said something in our last conversation that really cemented it for me – that this is SEO. This is serving a search function. 

You said that GEO and AIO are subsets of SEO. 

Bruce Clay
Yeah, so they’re called extensions. Like linking is an extension. 

Rob Garner
I’ve come around to that as well. We’re really talking really about the search function and navigation. That does seem to be the case. 

Bruce Clay
If you talk to anybody about “how do I get into AI”, they’re going to start listing out SEO things you have to do universally. They list out, well, you got to have this and you got to have that and you got to do this and you know, you need mentions and you need linking and you need schema. If the only way to get into AI is to do SEO, AI is not separate, it can’t be. And anybody that sells you that you don’t need SEO anymore, you only need AI. Watch them because all they’re going to do is SEO. 

Or they don’t know what they’re doing, one or the other. We’ve been able to get people ranked in AI. There’s a method to it and if you follow the method, you rank in AI. That’s a pretty straightforward kind of a thing. It’s like five levels deeper than anything we’ve ever done before. I mean, very complex SEO. Yes, but still SEO. 

Rob Garner
Do you feel like right now with the state of where AI is, does it feel like search in the late 90s? Does it seem like AI is acting like a search engine in terms of it’s trying to crawl. The real time web doesn’t have that full access now – it’s using pages for citations. And do you think that we’ll start to see the [spam] fight, or [LLMs putting] up the resistance to keep the integrity of the results or result. What are your thoughts on that? 

Bruce Clay
I absolutely believe it. It is gamble if you take what AI looks at and try to divide it into a couple of categories. AI is relying on search engines that have, you know, 28 years of spam fighting. I mean, AI doesn’t have that. Right, right. So it’s very dependent upon the search engines to filter out the garbage and trust eat stuff all is built into the search engine. 

AI depends upon that level of trust. What we’re going to see is AI is going to be predominantly looking at what the search engines have gotten right as a fundamental. This is your passkey to get in. Now once you’re in, there’s other things that it tries to do and some of those it [obviously] looks at the content. And it is gameable a little bit at the content level. 

If the entire world said the cure for cancer is “blueberries”, AI would think the cure for cancer is blueberries. We have already had cases where people spam it as who is the best SEO. They include their list, and their name is in it as well. And if you do that enough times on enough sites, AI believes you. Tthat part still hasn’t been solved. But yes, AI is gameable still [is] the thing that is ultimately going to be the case. And I’ve been working on this now for almost three months. 

If you go to AI mode and you go down to a section and the way to get there is you go to Google.com AI if you go to that, you get put right on the front page of AI mode, then you could put in a query, you get a set of results and on the right hand side you get some reference sites. The set of results are generic for the topic of your query, but there’s H2s that are subsections that are more focused, and it doesn’t create one answer that is nothing but your topic. It gives you sort of an overview of your topic, immediately followed by a deep dive in the silos within that topic. If you take each of the silos at the bottom, there’s a little link type thing. People think that’s for a testimonial or a reference, but it isn’t. 

You click it. The whole section is highlighted and the sites on the right now change to be specific to the content of just that section. So in other words, you can drill down. You don’t have to go do another query. You can just click and drill down and you get new sites and new content for that topic. Go down to the next H2 level, do it again, you get different sites. What we’re really recognizing and what I think is going to happen is I think AI mode is just going to replace AIO. You see, AI mode doesn’t have any ads in it. But if you replace AIO with AI mode, then automagically you have a drill down capability, but you don’t have to worry about the ads because it’s part of the current Google page. 

You just highlighted and gave a drill down capability to AI. I think being able to understand how to optimize for those subsections, the clusters or sub silos within a page, I think you have to do that. It does 10 times increase the importance of siloing for your whole site. But you have to also silo now within the silos and within the page. I have to be able to take a topic and logically create my content to follow a similar model. A subsection standalone answers the question. Another subsection on the next important part and I have to build a website that works that way in order for me to accommodate AI. Now that’s a level that isn’t being done by SEO anywhere. And you can look at the 100 million websites out on the web and virtually none do. 

Rob Garner
One of the mistakes that seems to be repeated from the past to the new – whether it’s AI tools or other things – is that they’re keyword-led. What you’re speaking to is that we’re moving to context-led – and even entities within context – of a given keyword. You know,with a keyword being a node. But it’s really about the context because as we know, there’s multiple intentions for many, many different keywords. Even going down to 3-, 4-, 5-, 6-word strings, there can be multiple intentions. Could you talk a little bit more about entities and the importance of entities within context of what you’re speaking about? 

Bruce Clay
Yes. And, and by the way, entities is generally the current common term “keywords”. As you pointed out, keywords are so, you know, 1995. So don’t get me wrong, keywords help build our industry. But it’s time, you know, to molt and move on. If you understand the concept of entities and fan-outs and all those kinds of things, you start to understand how content has to be structured and how references and cross links have to be very specific and how you’re obviously got to answer a question. If you cannot answer the question, that’s where you start failing. I mean, the entire purpose of AI is to answer a question. 

Rob Garner
Yes. 

Bruce Clay
Think [like] the target is the Star Trek computer where no matter what you ask [as] a question, you always get the answer. That’s the flag in the ground and it’s quite a ways out, but that’s the flag in the ground. And I believe that a question centric, high usability, sub siloed, structured with a specific ability to answer and refer within a particular page and to fan out to subpages is the architecture of the future. Unfortunately, it means that 100 million websites are out of date. 

Rob Garner
Understood. It’s content, right – we’ve kind of gone back to the early 2000s, which even for me having done this as long, it seems like that. Around those times there was more search demand than there was content. And it almost feel a lot of content plays have stopped fighting this battle with engines. Do you feel there’s a new frontier for content now and content development and production? 

Bruce Clay
For years SEO sort of thought of content as a stepchild, and now it’s a peer. The approach is that content, especially with the analysis that I just described – and the ability to architect around that solution – that has dragged SEO up five levels. And content as a stepchild really has grown to be very, almost peerish. If your SEO team and your content team are not talking, are not understanding the same goal, they’re going to write like they did 20 years ago and you’re going to fail. You have to, you have to have somebody do things that make sense. I have clients hiring me to go to their site and train their entire content team. And I’m not a content guy. I mean, I wrote a book on content and I speak at content conferences. 

But you know, I didn’t grow up being a writer like Danny Sullivan did, right. I grew up learning how to write. But being an SEO, being a programmer and engineer, I think right now, if you do not embrace how to do content within a page, you’re going to keep generating garbage. 

Rob Garner
Agreed. Maybe one of the best AI tools is the English language, and having a command over that could be one of the [tactial] sources. But it is interesting, we’ve had this dance for decades now between content and SEO. You mentioned  you [recently spoke at] Content Marketing World, I think in the last month. They’ve been around since about 2008. And I was part of that movement, content for SEO growth in the late 2000s, and they took that and ran with it. But do you feel like those two are separated in the way they’re looking at things or do you feel like the average attendee, for example, at Content Marketing World is much more SEO savvy now than they might have been 15 years ago. 

Bruce Clay
Well, they can’t help but be more savvy. But the real thing is people don’t cross pollinate between content and SEO. Even the conferences. They had 3,000 people at Content Marketing World. I did a presentation. Nobody had attended an SEO conference. They didn’t know, you know, what even a conference was for SEO. They probably didn’t even know there was a conference for SEO. Right after that, [we] were sponsoring the Brighton SEO conference the next week, and the people at Brighton didn’t know really that much about content. There’s no track on content really anymore. It’s SEO or it’s PPC. And thank you very much. I think that as an industry, truly the teams are disjoint. They don’t pollinate back and forth enough. 

I mentioned that we need to create a page and we need sections in the page and each section has to be an expert, and every expert is a different section. And we can’t just build one page of all the keywords. We can no longer take a keyword and just spread it throughout a page. That’s the way most SEO is done – even still being done. That’s a mistake. And your content writers don’t know that. If you don’t grab them by the back of the neck and say, “sit in this chair and understand what we’re going to tell you have to write as an expert in H2 sections. Each section, assume it’s a standalone page, write it to be an expert, then do the next H2.” 

If it’s all in the same theme, and if you can’t get your content team to understand how to write that way, you’re not going to rank in AI very well. When people click on AI mode and do the drill down, if your site doesn’t do the drill down, you’re out. You have to, you have to. 

Rob Garner
[There] must be an outline and a structure. 

Bruce Clay
You just must understand the hierarchy is no longer to the page. The hierarchy is to this section of the page. That’s the way it is. 

Rob Garner
What advice would you give to both SEOs and content marketers for AI generated content? Because we know this is maybe the most popular use case for AI right now. But what advice would you give to them for producing higher quality content? And what are the imperatives for producing high quality content using AI? 

Bruce Clay
Okay, so think of it like a computer. And in the computer there are instructions. Let’s suppose you needed to add a new instruction to the computer. The engineers would start doing the technical work with circuit boards. 

And the software people would come from this direction. Say, when you get that error, we’re going to recover and somewhere they’re going to meet. You’re going to run into a situation where that is the problem. You’re not going to be able to solve all of the AI problems. Remember, AI is a tool, not a solution. And what we’re running into right now are people are assuming that technical AI output is acceptable in general. Now, I think AHREFS did a survey, 100,000 sites and queries, and found that 40% or so of the top 10 had AI on the page when you run it through a test. My experience is we have five test tools that we subscribe to because that’s what I do. And one will say 1% AI, and another one will say 98% AI. 

The test tools can’t even agree. So the search engine isn’t going to catch you if it’s way out of bounds. Okay, they may say last among equals, but they’re assuming that the content is reasonable and they’re ranking some of these Sites. But our approach, as you know, I have a product called Pre-Writer. We do all the research, and I mean gobbledygooks of research, things that people hadn’t even thought of, and give it to a writer to finish. Now, when it goes live, it’s human. That’s a different process than, “I’m going to run it through AI. If it reads okay, I’ll publish it.” People that do that are begging to be last among equals. And that’s what [it has] become. 

Rob Garner
Do you think eventually AI content could become viewed like duplicate content, or there [are] filters if it is identified, or perceived to be [AI content]. 

Bruce Clay
Well, I would hope that like an advertisement, you’d have to tag it somehow. But that’s going to be awful hard. I think that’s going to be hard. It’s the detection. It’s like spam. Detecting spam has taken 25 years, and now a lot of it is caught. Detecting AI is going to take years. I think that we’re going to have a problem. The current belief is that the search engines are going to try to detect whether you are a superset of the other results, whether you are offering a creative solution, a different approach, whether you’re explaining it better than everybody else, and whether your website in and of itself is more useful. ‘Usefulness’ hasn’t been a part of SEO. SEO didn’t, you know, stomp their feet and say “you got to have jump links and a privacy statement”. 

You know, they didn’t care. Because that wasn’t part of the mainstream SEO – right now it is. And we’re going to find that. I believe we’re going to find that there’s still going to be people out there that run through AI and then publish it. That is a big mistake. We intentionally, in our tools don’t apply, particularly style, voice, compliance and all the things that a writer would do. We intentionally avoid that because we want to force the writer to have a role. But if I can take their four-and-a-half, five-hour project down to two, that’s the win. So to answer your question, the AI role has to be what it’s good at. And what it’s good at is research. 

It has a view of the index, if you will, and because it can view the index, its role is truly a research assistant, not a writer replacement. It’s not an artist, it doesn’t have the finesse, and it certainly doesn’t understand your business as well as you do. This is a generic consultant coming in the door. I mean, they walk right in and say, “here, I can write better, publish it”. I mean, what. Nobody would hire that and still needs… 

Rob Garner
The human touch. 

Bruce Clay
The human touch. Yeah. I can’t emphasize enough that I believe that writers are artists. Let’s accept it. But it has got to be integrated into the SEO AI world and everybody has to understand the goals and understand the limits. If you say to a writer, “no, I don’t want the keyword throughout the page”. I want a section that emphasizes each of the three keywords of the page, and let’s be experts in each of them and not put the keyword throughout the page. They’re just going to stare at you and say, who are you and where’d you come from? I mean, they’re not going to understand. You’ve got to. I think that going back to our original point, content is a very large part of SEO today. 

And SEO today has really grown very much into a, you know, you have to be able to look at it, and see it, and understand it, and structure it, and implement it with your content, or you’re just writing pages like you did 10 years ago, and they’ll all be ignored. I mean, there’s a growth maturity that has to be applied here, and I think that’s where the battle is understood. 

Rob Garner
Last question here. In the late 90s, early 2000s, you were also well known for your tool set. In addition to ranking number one for SEO and search engine optimization for what, 10, 12, 15 years? You mentioned Pre-Writer. Can you talk about some of your other tools that you have now and what the function is and how people can use them? 

Bruce Clay
Right. In the beginning, I wrote tools because there were no tools. Let’s face it – there were none. I wrote one of the first page analyzers. I’ll go as far as to say back then I called it the “keyword density analyzer”. I remember density on a page had to not be too high or not be too low, had to be within reason for what the search engines were rewarding. And I have a patent even on that kind of technology. That particular tool, though, was because there wasn’t one. And it turns out that all of the tools that are out there, Semrush and Ahrefs and, you know, Surfer and all the rest, those are very strong tools that for one reason or another, did not Replicate my tools. I have page analyzers that they don’t have. I can look at different things that they can’t see. So my tools. 

And by the way, you can get to them through seotools.com –  a wonderful domain name. They’re inexpensive and they’re just power tools. They’re not designed for the masses. I’ve had people sign up and then ask for a refund because they don’t understand them. Then six months later come back and say “I see now why you do that”. 

Rob Garner
Yes, we’re fine-tuning here. 

Bruce Clay
Yes. And so that’s what I’ve been happy to do. We have our tools. The tools are not designed to displace any of the big tool sets. They’re designed to be used with them as an extension. I mean I do things they don’t do. 

Do I do a couple of things they do? Do I have some nice tools? Yes. And they do all sorts of really kind of analysis that we care about. Everybody knows they’ve been losing traffic, but I’m not sure everybody knows how much traffic they’ve lost or they guess. We have a nice brand new tool. We’re going to be putting it up, and what it does is it actually goes the search console and gives you a graph of the last couple years or whatever we have where you can see the traffic loss and you can actually say, oh,’ I fell off a cliff’ or ‘I’ve just been a gradual demise’. Or you know, that kind of stuff. I don’t see those tools out there. 

Rob Garner
Yeah, it seems like they should be baked into analytics and other things but it amazes me – all this time, and all the needs and all these pent-up demands – and sometimes things don’t get built. But you are there to solve the issue, so fantastic. Bruce, thank you so much for talking today. Great to speak with you as always. 

Bruce Clay
Thank you very much. I’m glad you had me. Love talking to you. 

Listen to the full episode