Mike King: Our next speaker, Rand Fishkin. One of my personal goats. You know what I mean? Somebody who’s been meant a lot to me in my whole like career life and all that. Ran Fishkin is the founder of SparkToro, transforming how we understand audiences and the co-founder and CEO of SnackBar Studio. He’s the author of a fantastic book that I’ve learned a lot from, Lost and Founder, where he shares hard-won startup lessons and he’s a keynote speaker inspiring marketers to think bigger and smarter. Legendary marketer and SEO, Ian Lohrey introduced Rand to Dungeons and Dragons eight years ago and Rand loves it so much that he still plays every month. Presenting You Are Bigger Than SEO. Everybody make some noise for Rand Fishkin.

Rand Fishkin: Oh, man. We didn’t get enough intro music. I thought I had, like, 30 more seconds and I was gonna…do you have to pay if you use more than five seconds? You do? Dang. Wait. What’s the how how much are you allowed to use? I was kidding. Oh. He doesn’t know. You know, now I have to think about these things because I run a video game studio, and sometimes I want to use Italian folk songs, and some of them still have licensing requirements. 

Anyhow. Alright, friends. I am not in SEO. I haven’t been I haven’t been for over seven years now. It’s been a long, long time. And and still, I do come to a number of SEO events still. I get to see a lot of you folks around the world. I it’s because I I love people. I still care about this community and so many people in it, and I feel a debt of gratitude to this field, which gave me so, so much. I had my ups and downs with Moz, but, as many of you probably know, a few years ago gosh. When was that? 2021, after I left the board of directors, after I was completely out of the company, and had nothing to do with it, Moz sold to a private equity firm, I think a Canadian firm, and then they resold it internally, and now it’s owned by a different group. And I don’t know what’s going on with it. I think I only know Dr Pete there anymore. But, they sent me a weird check. Right? Like, I don’t know I don’t know if you know, but when you sell your company, like, you get a get a check-in the mail. I remember I was at at a cafe, like a a really, really terrible breakfast cafe in San Diego called the Point Break Cafe. I’m like, Keanu Reeves, Patrick Swayze, you can’t go wrong. Yeah, you can can go very wrong. It’s like where dock workers go to drink after their shift say, this is what you signed up for. Right? This is what you were expecting to hear? This is the whole talk. Geraldine and I had a rule. Like, if we ever won the lottery, you know, or something like that. Hey, Dan. Oh my god. Oh my god. Like, this is so cool. Wow. This is amazing. Hi. Hi. See, this is what I’m I’m telling you, people. You just you get to see your friends and meet your friends’ kids who you’ve never met before and heard so much about for years. So I’m I’m at this cafe in San Diego, and I get a a notification in my email that says, you have received a deposit larger than $100. It was a deposit so much larger than $100. Like, it was a it was an amount that looked stupid. I remember thinking, oh, no, the FDIC doesn’t insure my bank anymore. Like, I’m I’m finished. I need to move this money. I don’t have another bank account. What do I do? I only have a checking account at a credit union. But we had this deal, Geraldine and I had this deal, that if we if we ever won the lottery or, like, came into a huge amount of money or something, we would we would buy dinner for everybody at whatever restaurant we were at, like, know, that night or that week. Well, we’re at the Point Break Cafe. There’s about 30 absolutely sloshed dock workers, and a few folks from the veterans of foreign wars, like, club that’s nearby, and they’re having breakfast, which is a surprising amount of alcohol. I don’t recommend this, by way. Don’t get drunk at 8 AM or whatever it was. Anyway, we paid for everyone’s drinks. My favorite guy, this is my favorite guy. He goes up to the counter. There’s only one more woman working there. Her name is Nana, because of course. And and Nana says to him, honey, your breakfast is free this morning. And he’s like, wait, what? Why? He’s like, well, these folks are are buying. They came into some money. He’s like, no, for real? For real? And then he slaps down a five and he runs like the FBI is after him, like like, runs out the door like, okay, I’m just gonna take this one opportunity to not pay for my 17 beers. The whole check was like $350. It was not I kind of was like, oh, you know, I think we could have afforded that before the giant giant deposit. And then and then we did my absolute favorite thing ever. And this is this is the point of the story. We did my absolute favorite thing ever, which is we went to a nearby mall where we met up with Geraldine’s nieces, who have parents who both have doctorate degrees. They work in public health. They they help folks with with mental health challenges. And, of course, if you’ve anything about this field, requires many years of schooling and pays nothing. And, so they had told their girls who were both in high school that they would they would have to forego college, maybe do community college, take out some loans, whatever. And so we got to the mall and we were like, hey, who wants to go to college? And that was that was tied with my wedding for the best, best day of my life. Absolute best day of my life.

SEO has given me so much, so much. I I don’t think I’ll ever be able to repay it. And also, I wanted to do that for you. I want you to have that opportunity to, whatever it is, send your kids to college, send your, you know, if you’re like me and you’re shooting blanks, send your nieces to college. I don’t know if you know much about Judaism, but for like 4,000 years, we weren’t allowed to marry anybody outside the religion. And it’s not great for your gene pool. Like, yeah. It’s anyway, many health issues, which I’m happy to discuss with all of you. I’m sure since you’re in New York, you’re very familiar with Jewish American culture. Okay. Let’s let’s dive into this talk now that we have a quorum.

So I my questions for you are or your questions for me, I think, are why? Why, Rand? Why do I need to be bigger than SEO? Need is a strong word. I’m going to nudge you and encourage you, but I don’t want you to think that I am telling you that SEO is dead or over. I heard that for 20 years while I was in the field. It was bullshit every time. It’s bullshit now. You see it, you hear it, it’s wrong. It is changing. It is changing much more than I think it’s ever changed in all the years that I was in the field, and certainly the last few. Is very different. I want to talk about that. I want to talk about what’s really changing. Thank goodness we have Eli Goodman here in the audience. I am going be referencing data from his company, Datos, a tremendous amount because we work with them. Datos is sort of a big part of the back end of SparkToro, my audience research company, and it is a huge part of what I’m going to be showing you data from today. I’m happy to talk about that, and if I misspeak, Eli will correct me. He’ll be like, woah, woah, woah. That’s not how it is. And the third thing I want to do is make you indispensable. I want to make it so that you get to have the career long term that you want to have in marketing or entrepreneurship or SEO, whatever you want to do. That that’s that’s what I’m trying to help with today.

So let’s start with the first one. Why why do you it’s so comfy in this SEO cocoon. Remember remember when they first made us? They, I mean the search engines, like they first made us SEOs back in the mid-, late 90s. And when they made us, like, we were these we were these little robots that they put together, and our job was like, you know, well, what we what was our job? Like, what do we what do we do? We built links. That was I could have said we stuffed keywords, but I will tell you, I spent I spent a good five years building links. That’s all I did all day, every day. It was soul crushing work, for someone in their early twenties, and I I can’t recommend it. I’m kinda glad we don’t have to do that anymore. Not nearly as much anyway. But now I, you know, now I look around at the field at what we have, and I kinda ask the question like, do I think I could change these results with link building? Maybe. Maybe maybe a little bit, possibly. It’s probably not the most effective way to change them, but I could maybe blunt force it if that was the only tool in my toolbox. What about five years from now? You think you’ll be able to change these results with think building? I would not put money on that. Not saying it’s impossible, but it doesn’t doesn’t feel likely to me. How about this? Can you change these results with link building? Oh, also probably not. Probably even less so. This is likely the direction the other thing’s going in. Well, maybe you’re thinking, but Rand, you you minimized our purpose. Like, really what the purpose of an SEO was for 20 years was to get traffic, which is fair. I think that that’s pretty fair. And I still have I have I have a sinking, horrifying feeling about where that’s going to, and I bet all of you do, too. I I have seen some of the updates, right, on, like, LinkedIn and Blue Sky from SEO Week. I am not the only person to be talking about the end of traffic. We we know it’s going. If I were to perfectly optimize this result, if I were to make this the most SparkToro friendly thing it could possibly be, which, by the way, this is an insanely SparkToro friendly thing because of my colleague, Amanda, who basically wrote this audience research guide that she published in February, and now most of what you see in those results is coming from that and a whole bunch of other stuff that she and I have been talking about on the Internet, on podcasts and YouTube channels and webinars and at conferences and events and in every publication that we possibly can, which means large language models pick it up, which is great for us. Result is way better for us than it was five years ago. But five years from now, I don’t I still don’t think this is going to send us any traffic. It doesn’t send me any traffic today, and and that’s my result right there. The top one links to me. Like, pretty pretty optimized. And I I can guarantee you this is not going to send me any traffic.

So, SEO, let’s take the classic definition of of this field, right, is the practice of earning rankings and getting traffic from search tools of all kinds. I don’t want to limit to traditional search engines. I want to say anything you use for discovery. It could be Pinterest. It could be Reddit. It could be Google and Bing. It could be Perplexity and ChatGPT and Claude. And for a nice, comfortable 30 years, that profession grew dramatically because rankings impacted traffic, and traffic from search kept growing. And all of a sudden, I’m fudging a little bit, it really was 2023 that this started turning around that the impact that rankings had on traffic started to decline. I would say that was even declining, in decline from like the 2017, 2018 era to today because Google was answering so many queries without a click. Right? And and you’ve seen my research around this. But now, the traffic availability is flat or it’s falling. And I I don’t know how many of you watched Sundar Pichai and the the executive team from Google five days ago, four days ago, had their earnings report, right, the Google earnings report. Someone from JPMorgan on the call, one of the analysts, asked, hey, does this mean they were talking about the use of AI Overviews and Google was…the Google exec was describing how behavior has changed from AI Overviews. And the JP Morgan analyst asked the execs, hey, does this mean that clicks to the open web are falling? Because last year you said they were flat. Last year you said they weren’t falling, that Google was sending more traffic than ever out. I don’t know if you know this, but we Eli and I actually validated that with Datos’ clickstream panel last year. We were like, oh, look at that. Sundar is not lying. I mean, was lying about these other things when he spoke to Congress, but not this thing. This thing, he was telling the truth. It looks like Google actually did send out more traffic than they did the year prior. But now, now I don’t think that’s the case because you know how the Google exec responded?

Now is not the time to get into clicks and traffic. Now is not the time. Bro, you might have just said, oh, we’re throwing the open web right under the bus. Like, y’all are done. So, I think we have we have three options here. Right? We can change the definition of what SEO is. Like, we can try and expand it and basically say, yo, SEO is not about traffic. SEO is about influence. Right? SEO is about marketing broadly. Well, look, I think for you personally, you can change your definition. That’s not that hard. You want to change an industry’s perception of how people feel about a field and what they think it is for and what it does? Good luck. Let me tell you what I stopped doing in 2009, SEO consulting. How many emails do you think I get every week asking me to do SEO consulting? You want to change someone else’s perception of you after you’ve already built a name for yourself in the field, or after a name has already been associated with something? Man, forget about it. I was born in New Jersey. I am allowed to use this word. Oh, man. Wait. I thought I had a Kleenex in my pocket. Charlene, can I have a Kleenex? Alright. Second option, you can keep doing what you’ve been doing. Like I said at the start of this, I don’t want to give you the impression that all these jobs in the SEO field are going to disappear. That is not going to happen. Will the field be 15% or 20% smaller than it is today in terms of the number of people who actively have SEO in their job title or role or how many people are hiring for that or how many marketing agencies are hiring for that and describe themselves that way as a percent of the whole field in five years? Yeah, I think so. I think I think we probably I would expect that 15-20% shrinkage. That’s not that’s not the end of the world. Right? Plenty of people are still going to be doing this work. The field today, by the way oh, you’re amazing. Thank you. Also, could I request that you briefly mute the microphone? No. No. I’m fine. Thank you. Lousy bodies. I was being biological. All right. I did wear a mask on the plane, by the way. I’m like the only person in the US who still wears a mask on planes, but it’s gross. Like, you sit next to people and they’re just blah. Anyway, okay. 

Keep doing what you’ve been doing. That’s fine. And the third option is you can become something more. You can become something more. I’ll tell you what I’d like to do, Mike. What? Next year, if if you want to have me come to this event I know it’s weird to have a non-SEO person at an SEO week, but if you want to have me come to this event, I will walk you through how I basically used all of the SEO skills that I have to design and build and then fundraise for and then build the the video game. And we got this really cool screen. I was thinking we could just, like, set it up on a controller and you guys could play. That’s what I like to hear. When’s the official, like, wedding party? I’m still sore about that. So, I think among these options, you get to choose. I’m to talk now about what I think the future is going to be and what the present is. I’m going show a bunch of data, I’m going to walk through some classic Rand Fishkin-style. Here’s a bunch of charts and graphs. Here’s what it means. Here’s what you can take away from it. But these are your options, and I am not disparaging any one of these.

I’m not trying to tell you you can’t do SEO anymore. I’m trying to tell you you have choices. I want you to have a big life. I want you to have big opportunities. These are ways. These are ways to do that. Alright. Let’s talk about what’s really changing. Just AI stuff, right? No. When people ask me, you know, reporters call me all the time and they’re like, hey, you know, what’s changing in SEO? Well, I’m not in SEO. I don’t know why you keep asking me. Well, I do know why you keep asking me because there’s no every time someone publishes a list of top SEOs, I am still on it despite my best efforts. I can’t I I don’t know what I’m gonna do to get off of that. Alright. So 2025, bruh. My my nieces who are going to college have taught me that this is a gender neutral plural pronoun, like y’all, which I I very much appreciate. I thought English only had the one, but apparently, we have this one too. So, this I don’t know if you’ve watched human beings, but they don’t do it anymore. They’re, like, fully against it. No. We don’t we don’t click. We don’t like clicking. Click is bad. Swipe. Yeah. That’s what that’s 2025, baby. That’s where human beings are going.

When I talk to you about how all the platforms I’m not going to show you all the social platforms and content platforms and all that kind of stuff, I’m going to focus on Search since we’re at SEO Week. But when I talk about how all the platforms are taking away clicks, that is not just the platforms. It is also because we human beings don’t like to click. We don’t like it anymore. Like, we’re over it. We’re not into it. But, of course, if you are a big tech platform, it’s 2025, and so you’re like, woah, woah, woah. Send more traffic? Keep more traffic? Yeah. That’s the one. That’s that’s what I like to do. Stock bonuses, bruh. Right? Everyone I don’t I don’t know if you read the this was one of my favorite things. When, when the Google, lawsuit was happening, right, the Department of Justice is suing Google. They actually won the case. I don’t know if you saw this, which is quite exciting, and now there’s remedies potentially coming up. John, what’s up? But, hey, I bumped you on LinkedIn last night. Yeah. So so anyway, in that lawsuit, there was this deposition, right, and one of the documents that they got, that the Department of Justice got from Google’s internal servers was my favorite thing ever, which a Google executive writing to the other Google executives in email that said, hey, friends, what if we made search quality a little worse so that people would click on the paid search results because then our stock bonuses next quarter would look really nice. And I don’t know if you know anything about the housing prices in Menlo Park right now, but, like, stock bonuses, bruh. That that email I am paraphrasing slightly, but that email exists and is real and you can go read it on the Internet and it just chef’s kiss, all the, like, dark evil things you thought Google was doing. And they’re like, no, we don’t do that. They totally do it. It’s awesome. 

Alright. So let’s talk about cert.  Which side is it on? Oh, yeah, fun. Okay. So, you saw this big hullabaloo, right, when Google’s search market share is below 90% for the first time, there’s like, the press is going wild over it. Search engine land was only one of like fifty media outlets to cover it. This was a 1% drop, which I I don’t know, maybe 1.5%. If you look at like all over all time, Google’s market share, global market share, according to Stack Counters, has hovered between 89% and 92% forever. I don’t this doesn’t seem like a big change to me. It went right back up the next month. So Google but Google had to come out with a statement because at the what was that, the the New York Times event in December, I think it was, Sam Altman gets on stage from OpenAI. Right? Sam Altman from OpenAI, runs ChatGPT, gets on stage and he says, hey, ChatGPT is like this huge. I can’t remember the exact number. I have it in the presentation. I’ll show you in a sec. But he says like, hey, ChatGPT is is super huge, and I think Google felt some pressure to have to respond. And so in January, they come out with this statement that they haven’t come out with a statement like this since 2016, that they now process five trillion searches per year. One of my the most fascinating things when they when they talked about this four days ago, this number, they mentioned that a large portion of these, like somewhere in the range of 10%, are happening through Google Lens. There are people taking photos of, I don’t know, the tree in their backyard and being like, can I eat this? Not a great idea, but I’m glad they’re checking. It’s really helpful. Deaths from poisonous mushrooms are down, you know, a ton. It’s really nice. 

So along with Datos, Eli and I basically looked at the growth of Google searches to try and validate whether it is really true that Google is sending more traffic out and whether their Sundar statements from January are accurate. And here you go. So this is year over year growth. You see, well, you can look in any direction. I’ll show you over here. So in pink, right, is 2023, in blue is 2024, and then that gray line is showing us the percentage growth, and you can see that indeed all Google searches are up almost 22%. That is that is massive. Google hasn’t grown by 22% since, my god, probably like 2016, 2017, when he talked about how AI Overviews and AI in general is boosting Google’s search demand dramatically despite the press saying the opposite. Like, if you read, whatever, The New York Times or Bloomberg or CNN or any of these mainstream, they’re all like, AI is killing Google. Nah. That ain’t happening, bruh. Correct correct use of the term. I’m really I’m working on it here. Alright. That is four ChatGPTs of volume. Google grew by four ChatGPTs in 2024. Absolutely huge. If if you like seeing it in pie chart, that would be that would be what the pie chart looks like. It is Google, Google Images, maps, shopping, video. That’s pretty damn impressive. Danny Sullivan’s new thing. Right? Remember when he used to be one of us? The Google Web Google search, which is like just classic ten blue links, that’s getting a respectable 0.04% of all searches. Alright. But are they sending less traffic? It’s a good question too. So, Eli and I looked at this as well. I’m going to use Eli as a shorthand for Datos. I realize that Datos is much more than one person, and in fact, you probably didn’t do the polls yourself. But nevertheless. 

Alright. So 2024, this is mobile and desktop combined. Right? And we are looking Datos’ panel gives us what happens in a browser. So this is not things like Google Lens or inside the Google Maps application. Right? This is what happens inside the web browser, inside Chrome or Firefox or Edge or whatever. So 41.5% make a click. This gentleman in the pirate hat here, can you point to that 1% right next to your head? Okay. That’s an important one. You see he’s pointing that? Great. Great job. 1% going to paid clicks. I bet there’s a bunch of people in the room who are like, this data must be wrong. There’s no way this is correct because I know from my paid search accounts and I have tons of things that we regularly see 3% to 6%, sometimes more, on our paid search clicks. So how could it possibly be that only 1% go to paid search? Give me just a sec on that. 37% do nothing. Right? Like, perform their search and they are happy. They got their answer. How old is Tom Cruise? Way too old to be doing impossible missions eight times. I think those missions must be a little more possible than he thinks.

 21.4% perform another search, meaning they don’t click on anything from the search they perform, they just do another one. That’s what we call zero click searches. Right? And that number has been trending consistently up. I actually have the latest data in my inbox from Stanislav on Eli’s team. And I I we’re not publishing it quite yet because we’re still validating some stuff, but it is showing a direct increase, right, from like in the low fifties two years ago to in the mid-60s now. That’s just like the way it is. Right? Less and less traffic is being sent out per search. Search is growing fast enough to absorb that, but for every thousand US Google searches, 360 clicks make their way to the open web. So two thirds of all the searches end inside Google’s ecosystem, which I would suggest is still an SEO opportunity. I think all of that is still SEO. We compared, by the way, the EU to the US because you’ll remember that the EU has a bunch of strict legislation. Don’t know if any of you have traveled around the EU lately, but if you have and you go, like Ross has probably been in the EU, And if you have traveled around the EU and you tried to use Google Search, right, and you search for your hotel, you don’t get the Google Maps listing right at the top because that is illegal. Google cannot just put their stuff right at the top. They can’t put Google Flights right at the top. They can’t put Google Hotels at the top. So you get kind of a different experience. And yet, when you compare these, that legislation is making almost no difference. Sorry, EU Commission. It’s it’s not working what what what you’re trying. And this analysis was done in May and June of last year. So have those AI Overviews made things worse? Almost certainly they have. Here’s some data from Seer Interactive, our friends over there, that looks that looks bad. If you rely on organic Google traffic, you’re in trouble. Like, trouble. Right? I’m sure I’m not the first one to say this. This was this was the Seer estimate, right, that AI Overviews likely reduce organic by 20-40%. Google’s gonna have to grow fast to make up for that. 

So there’s this Bloomberg piece from just this past week about this where Google’s AI shift is is leaving websites feeling betrayed. Feeling betrayed. I’m sure all of you are feeling that same energy, hey, Google was this place where we got traffic and, like, that is ending. Also, if you scroll down in the article, the irony is just thick. Dear god, if I were Davey Alba or Julia Love, I’d be so pissed. Like, what just let them read the piece. You’ve got to put the takeaways from the AI summary above the thing. This is a piece about AI taking traffic from publishers. Have some decorum, man. Anyway, I have marked this up in Gen Z lingo. Alright. What are people actually using Google to do? Well, we we looked at this as well, and this is one of the remarkable things about a clickstream dataset, the size and scope of Datos’, which is I was able to take all the searches, right, like not all the searches. I think they sent me a few tens of millions of searches that happened across the panel. And then I was able to categorize those no, I’m sorry. It was billions of searches, but tens of millions of different search keywords that all these searches were for. And then I classified them. This is something AI is awesome at, by the way. Once you figure it out, I had to spend like an hour on the phone with Brittany Mueller to get all my AI classification systems going. But there you go. Branded keywords are about a third and and, sorry, a third of all keywords, but 44% of all searches. So almost almost half of all searches are for an existing brand. Like, people are not searching for, you know, high-tech home speakers. They’re looking for teenage engineering. I don’t know if you are familiar with this brand. It’s the coolest effing thing ever. I just got my first, like, teenage engineering device and it’s oh, they’re incredible. And I got it before the tariffs. Yes. Good timing. I actually when I saw the tariffs, was like, I’m ordering this. So there you go, 44% for specific brands. What about search intent? We can actually get to that too through classification. I hand classified a few thousand of these, which was, ugh, speaking of soul crushing work, so that I could validate how accurate the AI was at doing it. It’s like 96%. 96% is what I was able to get. So high, high percentage that we’re matching there. So search intent, you can see there, two thirds informational, about 20% commercial, Lots of navigational too. Right? 12%. But when you go by the percent of queries, right, the volume as opposed to just percent of queries, like you’re classifying keywords, but then you weight it by volume, you can see that navigational searches, like I just want to go to this website, is almost a third of all Google searches. 

That that’s pretty interesting there too. And then we classify them by topics. And you can see here, right, like arts and entertainment, which is how old is Paul Rudd or how many Mission Impossible movies can Tom Cruise possibly do. The 50% of these are what I would say are easily answerable topics inside an AI Overview or an instant answer or one of Google’s many systems, Google Maps, Google Hotels. And then we looked at the distribution of demand overall. And I find this fascinating because the top hundred queries, which are almost all navigational, have as much demand as the bottom 200,000 keywords. So it’s, I think it’s something something on the order of 75% of all searches are are in the top like 10,000 keywords. We we are moving away from a big long tail of Google, which was like the big promise of the open web, to a world that is much more consolidated. Yeah, there you go. 75% of all demand from the top 20% of search query terms. This looks like it’s increasing, by the way. If we go back in time, this looks to be growing. 

So here’s my takeaways. I think Google’s using these AI Overviews to grow fast, there’s no doubt about it. There’s not even a hint of the death of search. Like, anyone who says search is dying, they’re full of it. Anyone who says SEO is changing, they’re right. Like, absolutely. Zero click search is growing so fast and so much that I don’t unless Google has a bunch of more 20-30% growth years in it, we are going to see traffic from search engines fall and fall precipitously. Plus, almost half of searches already know what they want. You are not going to influence them by ranking number one. They already are searching for a brand, a website, a URL, something that they know they want. So I want to spend a couple minutes on AI tools because I know that’s of deep interest. And we have some amazing, incredible data here. This is basically just an ad for Datos, which because that’s I love it. Like, where where can you get this? Okay.

So what we what Datos looked at here is the percent of their panel, right, these are like millions of devices all around the world, and what percent visited any of these AI tools. And so you can see, wow, like 21% of all the panelists, of all these millions of devices, which are not professionals in budget industries, cross all demographic divides, visited ChatGPT at least once each month in Q4 of last year. That’s that’s huge. Other other, AI tools like Perplexity are also growing quite fast. I think Perplexity had the fastest 2024 growth of anything. They’re also one of the smallest ones, so speed of growth is related to how big you actually are. And then you can see kind of active users and comparison from SimilarWeb, which has a panel like Datos’, so you can get the sense that like, oh, okay, we’re all copacetic here. All the data agrees. But then, like, 99.8% you know who the 0.2%? It’s that one guy on LinkedIn who replies to every single post you put up and is like, Google is over, man. I only ever use Perplexity. That’s him. He’s right there. He’s the 0.2%. Everybody else uses AI tools for some stuff, uses search, Google Search, and Bing, traditional search. So I want to try and do this right because the press is, like, crazy around this topic, and it really bothers me. So thanks to some recent announcements and data, we can actually try and answer the question, what percent of all search happens in AI? I think this is a big question, and it’s a challenging one. So ChatGPT announced they had 300 million users during this NYT DealBook Summit, and he said users send a billion messages per day to ChatGPT. A billion messages. So that’s, you know, back and forths.

Google announced that they have five trillion searches per year. So we can, using data from a variety of sources, we see what is going on behind the scenes, like how many of these messages involve different new prompts versus old ones. And the average is eight messages per prompt according to the dataset. So a prompt, like a conversation, right? A conversation you go back and forth with ChatGPT, an average one is eight messages in length. 30% of prompts, as classified here, have search like intent. They’re informational, navigational, commercial, transactional. This big purple unknown is not actually unknown. It is very known. It’s just not search like. It’s things like, do my homework for me or make my Dungeons and Dragons character or hey, will you analyze this code and tell me if it will run-in Java? It’s that kind of stuff. Right? It’s the things that we use AI to do that we would not use search engines to do. So this is thirty 7.5 million searches per day that are happening inside ChatGPT. 37.5 million. That is huge. This is a ton of search happening inside ChatGPT. Oh, wait a minute. 5 trillion 360 so Google is 14 billion.  You guys want to see the graphs? Yeah, of course, because this is meaningless without the graph. Well, there you go. That is my god. Excel could barely make the one pixel height for ChatGPT there. That is 373 times as many searches happening in Google every day as are happening in ChatGPT. So, you know, is it growing? Yes. But ChatGPT is going to have to increase their growth rate exponentially to catch up with Google at all. Even in the next, like, 20 years, I don’t think they can. Google grew by four ChatGPTs last year. So if you were to market share, like make the little pie chart there, like Bing looks like a champ compared to ChatGPT. DuckDuckGo is three times as big. I don’t know why nobody’s talking about optimizing for DuckDuckGo if we’re all talking about how do I get into ChatGPT. It seems weird. You know? Like, here’s the data. It’s telling you what’s going on. 

Alright. My takeaway is AI tools are tiny. They send a 10,000th of the traffic that Google does. And comparing most of what people do, like 70% of what people do in AI tools to what people use Google Search for is folly. That overlap, also very thin. Does anybody still send traffic? We actually were able to validate this too. So there’s the web’s largest traffic referrers from last year. This was done in January of last year, so data’s a little old now, but 75% of all traffic referrals come from search. Remember? Remember, we were little robots who were built to get traffic. That’s what SEO is, or search marketing more broadly. 14% comes from social, 11% everything else on the Web. But this is not where people spend time. That is not where people figure out what they’re going to search for. You don’t if you go on the subway and you, like, do an inventory of what everybody is doing on their phone, you are going to find, like, you were going to be lucky to find 10 seconds of someone googling. Right? Like, they are doing other stuff. They’re doing this stuff. Search is sending the most clicks out, but it is only the fifth most trafficked category overall. And the time spent on search is minimal, very minimal. Social, news, productivity, content, video, that is where people are consuming the things and discovering the brands and the things that they want to search for. That’s where interest is generated. I sliced and diced this data to only look at referrals to the long tail, so excluding like the biggest websites, and that is bad news because the biggest websites are getting more and more share of traffic over time and the smaller ones, smaller, I say smaller, we’re we’re still talking about, you know, sites getting millions and tens of millions of sites. That’s falling.

LLMs, if they get adopted as search replacements, they’re going to make this so much worse. The way I think this is going to happen is through Google itself. Google is going to become the LLM that we all use. I don’t see ChatGPT and Perplexity and Gemini sorry, and Claude replacing Google. Okay. Traffic, you want it, it comes from Google. Every source of traffic on the web, however, is falling. Social too. Social is falling dramatically. And every single platform is stock bonuses, bruh. They are hoarding traffic for themselves, and why wouldn’t they? If you work there, you would, too. Right? You work at Reddit, how do I keep people on Reddit? Not, hey, let’s send more traffic to the open web. There’s really nice websites out there. No executive says that. 

All right. I think this kind of our future. I’ll spend a tiny bit of time on this. Look, I think it’s probably time that we need to build on rented land. I know. I know I have stood on stages like this. For 20 years, I did and said, don’t build your home on rented land. Everything should be your website and your email list. And now I am telling you, that guy is now wrong. He is now wrong. I am very salty about this. I am pissed that Elmo came and blew up not only my democracy, but also my Twitter account. Not you know, they’re not equivalent, but still, just I’m pissed. Right? Like, this this was a huge source of influence and traffic and referrals for me. And still, I’m gonna do it again. I’m gonna keep building on rented land even though it might go away because this has way greater potential reach than our website, and I know we’re reaching the right audience through it. And you know what? Marketers are not coming to SparkToro the way they came to Moz in 2005. People are not watching Whiteboard Fridays on Moz’s website. They are watching me do videos side by side, 100,000 of them a week, on LinkedIn and Reddit and Facebook and YouTube and all these other people’s platforms. This is all rented land. I’m going argue this is rented land too. And I’m still going to optimize for it. Right? I still want this. I want to be there, I just know I’m not going to get traffic from it. I treat this just like I treat YouTube or a LinkedIn post. I don’t need the traffic.

Your new job wouldn’t say our new job, I like saying our, but I also recognize that it’s quite fair for me to say I’m an SEO, is to change behavior, not increase rankings. I’ll talk about the second job, B, right here. So, if you focus on traffic, if traffic is your KPI, that’s what your boss tells you, you get traffic, that’s your job, don’t you tell me what Rand Fishkin said at SEO Week, you get me the traffic, then you’re stuck with search. There’s not channel diversity here. Right? We’re talking about maybe a little bit of social, but almost all search. Like, you can’t get it anywhere else. You remember this, this big story from Q1 this year? HubSpot has traffic apocalypse. Oh my god. They lost like 80% of their traffic from their blog. But this was the first thing I did right after I saw that. I was like, really? Let’s see how the stock’s doing. Damn. They just reported quarterly earnings. That’s that’s weird. Maybe what if traffic and revenue aren’t the same thing? If you want to win that argument with your boss, your team, your client, this is what you show them. Like, yo, traffic down 80%. This is going to happen to us. I guarantee this is going to happen to us. I don’t know if it’s going be 80%, but your traffic will fall. It’s going to happen for almost everyone. And still, you’re not going to be losing your customers. I wrote a whole post about this and a post about how to measure things that are not traffic, how to measure influence with sort of a correlation based dashboard on brand interest, engagement, and those kinds of things. You can check both of those out.

If you choose channel investments based on referral data, you will become Google’s fool. Because let me tell you exactly what happens. Datos can tell you exactly what happens right after people consume content in all these categories. You know what they do? They go to Google and search for your brand. I’m not saying SEO doesn’t deserve the credit for that, but maybe SEO doesn’t deserve the credit for that. Maybe that was somebody else. Right? Maybe somebody else did that. Maybe you should be doing that. Influence, not traffic. And the last thing I’m going to say, credit to Dana Di Tomaso who helped me visualize this. So what I’m showing here is where my audience is influenced. You know, people who consume traditional news media or search for keywords in Google or follow YouTube channels or social feeds or industry websites and blogs. Like, this is where my audience is influenced, in all these places. And then I have numbers around how many people are in these places, like subscribing to all the relevant YouTube channels, subscribing to Google Discover and Apple News Source. And this is where we allocate budget. I’m making up an allocated budget. SparkToro allocates almost no budget to anything. There’s three of us. This doesn’t match. This looks bad. Like, that looks terrible. And I would encourage you to take that concept to your boss, team, or client and go, wait a minute. Are the places where our audience is influenced the places where we’re putting time and energy and budget? Because if it’s not, that’s no good. That’s no good. Every single audience is different. This is a sneak peek at the new SparkToro. Again, this is Datos data that’s powering us behind the scenes here. But I’m looking at people with public relations in their bio, and you can see which social networks they use and how much and more or less than average Americans, and this is very different from I looked at SEOs, people who have SEO in their profile and which search and AI tools they use, and yes, y’all are big into AI tools. You see those huge bumps, ChatGPT, Claude, Perplexity? Massive spikes. SEOs flipping love these things. They visit them a lot more than average. Landscaping professionals, not so much.

Be careful, friends. Like, do not assume your behavior is your audience’s or customer’s behavior. Go research it. You don’t have to use SparkToro, like, go figure out where your audience pays attention and be present in those places. That is your new job. Thank you very much.