The Future of Search with iPullRank:A Recap Discussion of Google I/O 2025

Conducted May 22nd, 2025. Lightly edited by ChatGPT and Gemini. Tenderly reviewed by Garrett Sussman.

Garrett Sussman: Got the Wi-Fi code. Got the Wi-Fi code. Yo, what is up? Okay, we are here. We are locked and loaded. I am excited.

Thank you, everyone, for your patience. We’ve got a bunch of people who are waiting to talk to you here. Mike, you’re at Google I/O, man. I just want to hear you talk.

Set the scene and set the stage. How did it go down?

Mike King: Yeah, you walk up in there. First of all, Google is a remarkable company. You kind of forget it when you’re just outside and you’re just using the product, trying to navigate all its idiosyncrasies. But as an organization, they are so polished.

They are so on point. They’re doing really interesting things. Some of the products, when they talk about the things that are launching. I was at I/O in 2017, and I remember things they said they were launching that never came out.

Remember that thing where it was like it could make calls for you and stuff like that, and they were like, “Yeah, it’s going to be called Duo” or whatever?

That never happened.

You know what I mean?

Or like last year, when they announced that there were going to be ads for AI Overviews, and then that never came to fruition. However, at Google Marketing Live, which was the next day, they finally re-announced those ads, and they announced ads for AI Mode as well. So, you’re just getting clear indications of where things are going. It’s not so much about the speculation of it; they are showing you very clearly, “This is where we’re going as a company.”

And think about this, right?

Typically, the time between Google I/O and Google Marketing Live, I think it’s usually like a week or two, where they stagger it.

Instead, they had them overlapping.

I think that’s a function of them knowing that with the things they announced, there was going to be a drop in stock price. Then the next day, they announced all the things on how they’re going to monetize things, and then the stock price rebounds. You know, seeing that, I kind of wanted to just buy a bunch of Google stock on I/O day, but you know, I don’t want to be accused of insider trading or anything like that. I’m not trying to hang out with Martha Stewart in that way.

But yeah, when they did the keynote, there was an audible sense of astonishment from the crowd about some of the things they were showing, like the Android XR stuff, specifically.

With all the announcements in Search, I think the people in our space weren’t surprised.

In fact, you know, like a week ago, or actually, no, SEO Week, when I was talking about how everyone was saying like, “Oh, it’s just SEO.” And I was like, “Hey, yes, there is a high overlap between the AI Overviews and the classic organic SERP right now. But once they introduce things like memory and personalization, that’s not going to be true. Or like MCP and A2A, that’s not going to be true.” And literally, literally every single one of those things was announced as a part of AI Mode on that day.

Right? Like you got personal context, which is their form of personalization, and to some degree, memory across the Google ecosystem about you and your stuff.

And then you’ve got, you know, Deep Search, which isn’t rolled out yet, but it sounds like it’s an even deeper version of the whole deep research paradigm.

And then what’s the other thing that they announced? I can’t remember. Whatever it is, it’s like an aspect of AI Mode that’s just going to make it a user-to-results sort of experience.

So it all begs the question, where do we go in search, right? Like, what is it that we do? What do we have the ability to manipulate to drive a result? And I had some conversations with some Google engineers, and I explicitly asked that question. I was like, “Yeah, with all this reasoning stuff, all this personal context stuff, like where do you see our role aside from creating great content?” And he corrected me on the “great content.” He specifically was like, “Great and unique content.” And other than that, he’s not sure. He doesn’t know what we do. And, to some degree, I think that’s fine, right? Because we have looked to Google for too long to define what it is that we do.

And then there are those of us who have uncovered the nuances of how whatever they’re trying to do works, and then we do things accordingly.

So it’s really exciting for us. It’s a really exciting time for us. It’s not going to be exciting for people who can’t really cross this chasm and be creative and do the experimentation to figure out how we can manipulate these things. But for the people that are driven by the curiosity that’s created by, you know, effectively adding search algorithms, this is going to be the craziest and most fun time that we’ve ever had. Because, you know, the technology is being made right in front of us.

There are so many open-source versions of this. Google has been putting out so much information about how stuff works. And they’re also giving us basically the ability to use a version of the same tech that they’re using.

So, in Gemini 2.5 Pro, they say, “Yeah, there’s a custom version of this that’s used in Google Search.” Well, yeah, of course, we don’t get the exact custom version, but we can get as close as we can by using what’s publicly available and replicating a lot of what they’re doing.

From my perspective, what we’ve learned is that this whole query fan-out paradigm is driving all of this. And there’s no visibility into what those sub-queries are. And it could be, you know, in some cases, we’re talking about this Deep Search thing; it could be hundreds of searches that are happening in the background. But nevertheless, I think that’s where we need to start.

We need to kind of reverse engineer what those sub-queries are that Google is considering, and then look at search as a more matrixed idea. So for any given query that you’re putting in, let’s just say for simplicity’s sake, that there are 10 queries that are happening in the background that they’re pulling results from and then using chunks from those results to compose whatever that response is.

So that’s something that, to some degree, we already know how to do. If you’re doing AI Overviews, you should already be doing this stuff because query fan-out is being used there. And I think that that was a key thing that we didn’t understand when AI Overviews first rolled out, because a lot of us, including myself, were saying like, “Oh, this thing that ranks, you know, number 50, is being considered for this AI Overview.” And that may not have been what’s happening. What’s happening is that the thing that ranks number 50 for your core keyword might rank number two for one of these sub-queries. So it then pulls from that page.

So I think the first place to start is really figuring out what are those sub-queries, what ranks for those sub-queries, and then how do we better kind of create this surround sound situation where no matter what, Google runs into our page over and over and over as it’s considering what information to use to construct these AI surfaces. So, you know, I think that is where I talk about things like Relevance Engineering, which is a bigger thing than SEO, because this is going to require actual engineering to build these ideas.

And I came away with a feeling of kind of despair the night of the first day of I/O because I was like, what am I going to do? Go log into one of these SEO tools and try to do any of this? Because there’s such a gap now. The gap is tremendous between what Google is doing and what we have, the information, the data, and the tools to do something about.

When I talk about this being like a matrix thing, that means that you have to look at not only what’s happening with your core keywords, so like the top 20, 100, whatever it is, pages for your core keyword, and then extract features from those and get an understanding. Because then you have to do that same thing across 10, 20 keywords. What tool is going to allow us to do that? There’s none, right?

There’s no tool that is set up to do something like that. The closest thing that I can think of was Searchmetrics’ Content Experience because they were already looking across the topic cluster for information to determine what relevance is.

We need a tool that is like that, but a bit more robust, so that you’re thinking about optimizations on multiple pages for the same keyword. And so that’s what I mean.

If you want to do things that align with what Google is doing, you have to build those things. And you have to understand content strategy. You have to understand IR, AI, UX, and also ultimately, digital PR, because you want these messages to live in a variety of places across the web to reinforce what you’re saying.

So I don’t think that just going back to doing TFIDF work and whatever obsolete thing that SEO tools provide is going to allow us to have visibility on these surfaces in the future. So it’s not just SEO because SEO is not enough.

But anyway, yeah, so lots of things that were rolled out. I think for me, personally, I’m always searching for the nuances in how things work. And so what I do is I go around and collect as much information as possible and then find where’s the Venn diagram between what everyone is telling me. And again, the query fan-out thing is the thing that ties all of this together that makes it all work.

But I sat in on a discussion. It was something like “Under the Hood of Gemini” or something like that. And Liz Reid, the head of search, VP head of search, was on that panel.

You know me, I generally think panels are a waste of time. I’ve never once learned something significant from a panel. But she did say one thing that really stood out to me is she was saying that for AI Overviews, the query fan-out is simplified, whereas for AI Mode, it’s more robust.

So what that tells me is that they’re looking at fewer queries for AI Overviews and perhaps fewer documents for AI Overviews because it’s all about speed. She talked a lot about for something like Deep Search, someone is going to be fine with waiting like five minutes for a response. Whereas when you’re doing a search, they said something like if it took more than 1.2 seconds, users will just bail.

And this is something that I talked about in my post from almost two years ago. I forget what it’s called, like “How Retrieval Augmented Generation Is Our Future.” And I was saying how back then, when AI Overviews first rolled out, they were taking up to like 30 seconds to generate a response. And of course, a user is not going to wait for that, especially for this novel thing that you don’t even know is going to be valuable.

But now it’s almost instantaneous, right?

And so they’re at that point where they’re looking at that latency on that millisecond level. And so they’ve got to have tradeoffs between how much can be considered versus the speed of the experience.

So I suspect that AI Overviews is always going to be like a limited subset of queries that it’s looking at, whereas AI Mode can probably look at like, you know, 20, 30 queries and like 200 documents. And it shows you, as AI Mode is loading, it’s like, “Okay, ran 15 queries. Okay, looked at 3,000 documents,” or something like that.

And so I also, they had this thing called the AI Sandbox, which is just like this indoor experience with all these different things. Like one of the things they had was you take a foul shot, and then it rates your form and all that. And it gives you feedback on what to do. I was so embarrassed because I missed like five foul shots. But it was like, “Yeah, your form is perfect.” And I was like, “All right, great. You know, let me just make a shot then.”

Garrett Sussman: But anyway, I mean, Britney Muller shared that, and it calls out how much of Google I/O is curated because, you know, we saw the same thing with OpenAI and their demos. And then people went back, and they’re like, “Oh, this is all like carefully planned technology.” I’m curious, though, because you mentioned AI Overviews. Did you get an indication? Do you think AI Overviews are here to stay and they’re just going to be built out, or because of costs, AI Mode is the plan to replace that?

Mike King: So my feeling is that AI Mode will ultimately be the primary version of search.

I suspect that what they’ll do is, down the line, I’m not saying soon, I think probably a year from now, at the next I/O, they’re going to be like, “You know, we’re just going to switch.” Because, one, Sundar was saying how I/O was like the best search launch in 10 years, which for people that live in this world, we’re like, “What? This is trash.”

But what they’re saying is that what they’re seeing on their internal metrics is that things are better. And when they talk about the growth in searches and so on, that makes sense to me because they’re not measuring this in the qualitative way that we are. They’re measuring it in a quantitative way. And the way I always describe it is like, you know, if you order from a flower delivery company like an FTD or something like that, and the flowers are delivered late to your mom, yeah, that sucks for you. But they’re like, “Hey, we’ve delivered 98% on-time delivery for all the flowers across our ecosystem, and that’s up 10% this year.” So to them, they’re doing great.

And so when Google is measuring things like their IS score, they’re looking at the volume of queries, they’re looking at the volume of query reformulation, they’re looking at how often are users having to elongate their query or switch their query because they weren’t finding what they’re looking for. Like all their measures are saying that things are better. So from their perspective, I get why they feel that way and why they’re going in this direction.

There’s just the reality of the competition, right? Even though we all know from the data that OpenAI is infinitesimal compared to what Google is doing, the conversation is about how OpenAI is coming for them and how they’re beating them.

Then there are these aspects of these, like you know, stock counters saying that Google is going down. So all of that, plus all of the other threats, and like the antitrust stuff, Google has to go in this direction. That is just the reality of it.

So those of us who are hoping for things to go back to 10 blue rings, you’re going to have a really hard time. Those of us who are expecting traffic to come back to where it was are going to have a really hard time. And what was kind of fascinating to me was that, you know, in discussions with engineers, they seem to be shocked when we told them that, you know, clicks were universally down for all of us, right? Because there are a number of us who had spoken to a few different engineers, and, you know, they were asking questions about our reaction to the launches and things that we’ve been seeing as a result. And also, we were explaining to them like, “Hey, yes, we fundamentally understand like the goal here is to drive qualified users that do things on our website. It’s not necessarily about volume.”

But the reality is that traffic has been the measure in so many ways for so long that, you know, we have had to recalibrate with our clients what this means. And what would have been valuable was some sort of air cover from Google.

And this is what I said at, you know, Search Central in New York as well, but I reiterated it for some of these engineers. And I also leaned into this idea that like, “Look, this relationship between our community and you guys is a symbiotic relationship, but it’s too much like one-way stuff happening.”

You know what I mean? And I was using the example of structured data, of the secure protocol, Core Web Vitals, where they effectively used us to make something happen that was only to the benefit of Google. And now you kind of just have us twisting in the wind when you guys make this dramatic change, and now we’re all struggling to get resources to make things happen. So, you know, that didn’t seem to land on deaf ears. I think they heard us when I said that.

And, you know, we were also just kind of talking about how Google Search Console is not enough, right?

And I’m actually working on a blog post right now, and the more I think about it, the more I think it should be more of an industry study, but it’s basically about how I would have designed Google Search Console, like all the features that it should have within reason, right? But irrespective of that, Google Search Console is not prepared for AI Overviews or AI Mode at all. Do they care? I mean, yes and no. Like they care in that they want to have some sort of data source to provide to us. But someone explicitly said like any engineer that we put on GSC is being taken off of growing Search. And I get that. But because they’re like, “Oh, it’s a free product.” I was like, “Yeah, well, Google Analytics is also a free product.” You know what I mean? Like, there are basic features in GA that should also be in GSC. Like the fact that, did we just get annotations in GSC? GSC has been around for what, like 15 years or more? I can’t remember at this point.

Garrett Sussman: Well, there’s also the call out, too, that we’re getting the queries for ads. You know, there’s always that, you know, the way that the ads are treated and the way that organic is treated, right?

Mike King: And that hadn’t been announced when we were having these conversations because these conversations are on day one, and then the GML stuff came out on day two. So, you know, when we were discussing this, folks were saying like, “Oh, yeah, it’s so great that we’ve anonymized data. You know, it’s an important thing for privacy, yada, yada, yada.” And it’s like that argument holds water when you can say that in isolation. But then when you look at the context of the fact that the people that are paying for it are getting that data, it just breaks that argument entirely. Like, “Oh, so privacy doesn’t matter when money is exchanged? Like what are we talking about?” And, you know, like I…

Garrett Sussman: So let me ask you this, in terms of like short-term strategy, like the lens that we’re looking at this through, because we all know that this is kind of the beginning of a lot of changes. And at some point, search is probably not going to be a thing. It’ll just be kind of converged with the conversation platforms with like a general assistant. Are you looking at it through a short-term lens of what brands should do now versus once AI Mode, like say in a year, does become the default? Or are you thinking now and kind of future state of content strategy?

Mike King: Yeah, I think you need to live in the reality but also prepare for the future. So, you know, live for the moment but plan for the minute. So, yeah, I think right now, it’s really about what do we do as it relates to AIOs. And those are the things that I’ve been talking about, you know, for the last two years, basically, right? So structuring the content in semantic chunks and being explicit using semantic triples. And it’s interesting because there were two sets of guidance that came out yesterday, one from Google and then another set from Bing. Google’s was more of the same, right? It’s like, “Make great content, don’t make commodity content, use structured data.” And then the last one is like, “Oh, don’t measure traffic.” Like, dog, really? Like, that’s the best you’re going to give us after all this? Like you’re causing sites to lose anywhere from like 20 to 60% of their traffic.

And that’s all you have to say. And then Bing came out with some stuff, which said some of the same things. But also, I think it was like the last part of it, it talked about, you know, using header tags, keeping things tightly coupled as far as how you describe them and things like that. Really the things that I’ve been saying about what I just said a few minutes ago. So, yeah, so you need to figure out how to both tactically and strategically go after that. And I think the strategy side is where our space struggles the most.

You know, because strategically, what we need to be thinking about is like, search is no longer just a performance channel. And in fact, it never really was. Like it’s always been a brand channel. But we have discounted the brand side of it. And so with traffic eroding, you need to think about what is this channel actually doing?

And so an example that I’ve been using is like, if someone searches for, you know, “best basketball shoes for flat feet,” well, if Nike appears for that, that is a good thing. Your brand is now appearing for that non-branded search because a user doesn’t have to click through. They can just go to the store. They’re like, “Okay, well, yeah, Nike’s here. It’s the Air Shock, whatever.” You know what I mean? They just go to the store and get it. And so that is no different from a billboard. That is no different from a display ad. Those are all things of value in other channels. But in search, we don’t value that. We say like, “Oh, that’s just an impression. Who cares? We can’t track that down to a conversion.”

Conversions will still exist in this lens, just like you can still get conversions from display too. It’s just not the primary thing that that channel does. So yes, continue to measure those things, continue to go for those things, but expect that those numbers are going to come down because the user behavior is changing due to the changes in these channels.

And that’s what I’m saying. Like, strategically, SEO doesn’t have that conversation. All SEO does is think tactically and say, “Oh, traffic is going down. How do I get that traffic back?” The traffic’s not coming back. I mean, what you can do is focus more on the mid-funnel queries and the lower-funnel queries because those are going to be places where clicks will still likely happen because effectively Google is giving more information in the search so a user is more educated without clicking through. But they’re really looking to click through when it’s time to get deeper information or to take an action. So you do want to be in those positions to get more of that. But like this upper funnel stuff is not going to drive you much traffic, or at least it’s going to drive you fractions of the traffic that it historically did.

Garrett Sussman: Do people still need to do it? Like, in the eyes of, is it purely brand plays in the way that, like, kind of Google, do you think will kind of reward just the more in-real-life authoritative brands? Or for the sense of like machine learning and topical authority, do you still need to have that top-of-funnel content on your site? Or is Google smarter and doesn’t really care if—

Mike King: So if you are only making top-of-funnel content for search, again, you weren’t thinking about this strategically to begin with. You were just like rinsing and repeating the playbook of search from time immemorial, right? And on the IRL side of it for brands, let’s think about what a brand is, or at least the signals that a brand… And let’s just simplify this for the purposes of a 20-minute discussion. If we’re talking about the ABC stuff that Google was talking about, so anchors, body copy, and clicks, all of those things are better or are created by brands, right? Like what naturally gets a lot of links with targeted anchor text?

A brand, because you recognize this brand, you’re going to link to the brand because you trust the brand. Like that is the in-real-life thing that happens by being a brand. Who has the ability to invest in more robust body content? A brand, because they have money, they have resources, they have people they can put on them. And then who’s going to get more clicks? The brand that you recognize.

So it’s not necessarily that a brand is this thing that is just defined a certain way. It’s that the signals that Google looks at are best impacted by being a brand. So I would say that it’s not about SEO taking over the brand budget and doing all the branding work and so on. It’s about us leaning into doing things that brands do so that we have more of a footprint. And then that footprint is reflected in those scoring functions that Google is looking at and reflected in the features that Google is collecting about websites, about entities, about authors, and so on.

So that you are sending all the signals that they are actually looking for.

Garrett Sussman: So a question coming in from Usman, can you speak a little more to the brand channel versus the performance channel on like a tactical level? He wants to know, do you think impressions might be the new clicks to measure, or is it kind of, to what extent is a vanity metric versus an actually valuable metric, or what metrics are you looking at? How would you define that to the C-suite?

Mike King: Yeah, I wouldn’t call that a vanity metric because what we’re after in marketing is what? Attention. We want people’s attention. And so when someone types in a query and they’re going through this big AI Overview, if your brand is mentioned there, that is you getting their attention. It’s really up to how your message is presented and whether that attention turns into something else. And that’s ultimately up to you. So as far as what we measure, I do think that there needs to be like kind of brand-level mention measurement for SEO.

And what I mean by that is like, again, if you are, if your brand is showing up for a non-branded term, that is something valuable. So you should be measuring like for a given query, yes or no, is our brand there in AI Overview? And then if so, how many impressions are we getting on that query wherein we are being shown? Like that is something of value.

So I think our reporting should be reflecting that. Like I think that if your reporting is just about the clicks and the performance and so on, you also need to have a section in your reporting that is just about the brand impact or the brand visibility. Also, brand sentiment, just because if you’re showing up and it’s like a bad message, you want to know that. So yeah, I think that we need to be…

Also, I think that there’s kind of this misconception that just because it’s your brand that you’re going to be number one. That’s not always true. So let’s think for example, like a company like American Express. There are instances where NerdWallet will beat American Express on some of their branded terms because they’ve got like comparison content or they’ve written like a better article on the American Express Platinum Card or things like that.

So in some cases, your brand term is not this like obvious slam dunk that you think it is because there are other signals that may be overpowering. So I think brand measurement needs to be really deeply considered for SEO. And once we’re able to explain what that value is, I think that’s going to unlock more budget for SEO because brand gets more budget than performance does. Like that’s just a fact. And I think what also is going to happen because it’s happening in parallel with ChatGPT. ChatGPT does not drive a ton of referral traffic. So people are looking at it like it’s a brand channel rather than a performance channel. And being that Google is effectively becoming more like ChatGPT, the C-level people are going to start to reconsider how that channel needs to be looked at.

Garrett Sussman: You talk a lot about products. How do you feel about media publishers, like that are just where content is the business model?

Mike King: Yeah. It’s a wrap. It’s a wrap. I don’t know what to tell them. And I brought that up in the meeting. I was like, “All this stuff that we’re talking about with conversions is great because we are seeing that conversions are either growing up or I mean, going up or maintaining the same level with less traffic.” But that doesn’t work for a publishing model. It doesn’t work for an affiliate marketing model. And there’s no real answer for that. You know what I mean? Someone said something to the effect of like, “Oh, well, the quality of the impression is better.” But what I said was like, “Well, what about AdSense? You guys aren’t paying someone based on the quality of the impression. You’re paying them based on the number of impressions.” So unless the advertising models change, those sorts of companies are out of luck. I don’t really know what to say.

Garrett Sussman: What do you think happens? Because I’ve been thinking about this a lot with the affiliate space. You have the review space. You have like everyone who writes a middle/bottom of the funnel comparison page has an agenda, right? You have people writing about themselves saying they’re number one. You have aggregate sites like kind of just getting user reviews. We have the official critics. Do you think we’ll ever see an unbiased kind of recommendation engine from these search engines?

Mike King: You’ve also got SEO companies that are listing the top 10 SEO companies and putting themselves on it.

Garrett Sussman: Keep your eyes peeled. I’ll have that article actually out later this week. I’m not even joking. We’re doing an experiment. But yes, continue.

Mike King:  I mean, bias is a whole other can of worms. Everything is inherently biased. Britney Muller also brought up a great point about how Google used to present information in a simple way.

Now, they interpret the information before they put it to the public. So I don’t know that bias can ever be removed from anything. There’s always an inherent bias from whoever writes about whatever reason.

I’ve also always likened a lot of this review stuff to Nintendo Power. I don’t know if you remember that magazine that Nintendo used to publish, and they would review their own games. Are they ever going to say that their games suck? No, of course not.

So I think that’s just the nature of the internet. There’s not going to be any universal third parties unless we’re talking about Consumer Reports; they’re technically a nonprofit organization. So, what’s the incentive for an organization like that to exist where they’re not biased? So I don’t think that’s something that we’ll ever be able to overcome, irrespective of the medium.

Garrett Sussman:  To that point, I’m glad that you mentioned the bias because that was the big thing about the personal context. That’s what I talked about in SEO Week. What do you think are the actual implications of personal context and a more dynamic search result for SEOs not to be able to track or be able to target? How are you looking at it?

Mike King:  Yeah, rank tracking is a silly idea moving forward. Everyone has something that’s different. Because the way rank tracking generally works is you get a clean machine that’s never been on the internet before, and the first thing that it does is it performs a search, or in this case, triggers a conversation or whatever. And so there is no context. And what you’re effectively measuring is what would it be if there were no any other aspects that are being considered?

So you’re thinking about a sterile environment for ranking, and that no longer exists. And to some degree, it was already a fraudulent idea, and that’s why the average position is what Google gives you from GSC.

So I think the measurement of this is going to be tricky.

I think that what Profound does is probably the smartest way to do it, where it’s like, “Okay, are you getting mentions and citations, and how much is that fluctuating over time?” But again, even that is tricky because with personal context, I could see a dramatically different thing than someone sitting next to me.

So yeah, again, this is where it’s going to be challenging to figure out where we play in this because I think the first thought that the spammers of us is like, “Cool, I can just email you a million times about my brand, and then I’m going to appear in your personal context,” or, “I can forcefully share Google Docs with you and other files, so I’m in your personal context,” or any of these sort of things that are obvious hacks for how you can get into someone’s Google ecosystem.

But really, this all goes back to being a brand. If you want to be visible in these things, the scalable way to do it is to be the brand that someone wants to subscribe to the mailing list, excuse me, or be the brand that someone wants to download eBooks from and have them be in their Google Drive and all that sort of stuff because anything else is a short-lived hack that they can patch in, and then no longer is an issue here.

So yeah, I think it’s going to be a challenge, and it’s going to be a very dynamic one.

But on the other side of that, I think all of this is great for the user. I would love… because we’ve talked about the idea of the future of search being Her, like the movie Her, and OpenAI had the first version of just being able to talk to the thing, but now Google has literally replicated everything that Her could do in that movie.

Like, if you remember when he first boots up the platform or whatever, it goes through his emails, and it’s like, “Oh, I can delete all these and I can keep this, and I’ve learned all this information about you.” Literally, that is what personal context is. So I think it’s really exciting to see what the use cases are.

They also demonstrated ideas where it could buy stuff for you, it can track prices, and then when the price hits a certain level, it knows to buy it for you, and things like that.

So there are a lot of really cool implications of this technology, but as far as where we play, aside from having the big picture impact of being a brand, it’s to be determined what we can actually influence here. And I’m excited to figure that out. I’m not worried about it. It’s just like, “Okay, let’s dig into this and figure it out.” And again, I don’t think it’s going to be through the lens of content links and 301 redirects that’s going to get us there. It’s going to require actual relevance engineering to do it.

And that’s where I go to as well in terms of the combination of all things. Right now, we’re talking very much of the Google ecosystem, but you’ve got to think ahead. At some point, all of your transactions are data, all of your social media. At what point will Google or OpenAI include all of that? Because it already has our location, it already has our search history, it has a lot of things we’re doing online, but there are still big pieces, I feel like, that are missing that will eventually kind of fill in over time.

Garrett Sussman: We’re almost at the end of time. Does anybody have any last questions that they want to get to Mike before we bounce? Because—

Mike King: He said, “Are we holding our highest value stuff back, or should we be leaning into it to get a lot of references by releasing it?”

 I mean, I think that’s a strategic decision, right? I don’t think it should be a blanket in either direction. But I think from the perspective of competition, especially as we’re thinking about other channels like ChatGPT and things like that, I think it does make sense to hold back certain components of it.

Because one thing that we didn’t talk about is that Google is going to be moving, visualizing data directly in the search. The data visualization was one of the key things that we’ve continued to lean into because it’s like, “Oh, they can’t replicate that in the search.” Well, they’re about to.

You might want to make it so they can’t access your data and put it directly in the search, so that there is a reason for a user to continue to click through.

But as it relates to ChatGPT, I’ve been saying that cloaking is actually a viable thing in those cases. And before you start burning me at the stake, that shit, black hat is not a thing on that side.

Black hat is a function of Google calling it black hat.

Google saying, “Hey, these are our guidelines. You can’t do this to be in our website.” But ChatGPT does not have those guidelines. So it’s not black hat. It’s really—

Garrett Sussman: Well, that’s what I mean. It’s short-term, long-term, right? It’s like, how are you thinking about the consequences of that long-term? Because at some point, it may or may not be kind of—

Mike King: I mean, think about it. Because one, right now, they are not caching anything. They’re still crawling in real time. And there are all these reasons. They’re not searching.

Right. I mean, and they’re not even crawling. They are requesting pages. So their paradigm is different. And because they’re not indexing, what are you going to say? “You can’t have something on your website?”

Who are you to tell me that? You know what I mean?

So I think, specifically as it relates to competition, it’s too easy right now for your competitor to just grab your content and then replicate it using ChatGPT.

So, whatever components of your content that are unique, you may want to block those using JavaScript or render them using JavaScript because you can’t see them. So it isn’t as easy for your competitor to take your unique content and then have it on their site.

Garrett Sussman: It’s so unfair. There’s so much to talk about. I know we’re almost at time.

We didn’t even talk about your whole thing about content needing to up their game. Just seeing some of the Veo 3 tools that are coming out, and the capabilities and the deep research, so much is done for you. And it’s still high-quality content.

Not to call him out, but I saw Tim Suolo make a comment about how he has an entire team to do all these podcasts. He wouldn’t be able to do this with AI and repurpose all his content. I’m like, Tim, give it. Give it.

Mike King: Yeah, I mean, he had a quality standard in his mind. No, you could take an Ahrefs blog post, because they’re pretty long, throw it into Notebook LM, and then throw it into Veo and be like, “Okay, now make a visual podcast.” So I have no patience or tolerance for bad content anymore. There is no excuse. There are too many free or cheap tools where you can make a whole movie if you want to.

And the fact that most of my feeds are still the same corny memes used over and over, or bad one-liners from people who are not funny, is preposterous to me.

Garrett Sussman: Dude, I’m not stopping my bad one-liners. So you’re going to have to deal with it as long as that works for you. No.

Mike King:  So, the bad one-liners, I want to see an animation to go with it. There’s no reason you can’t make one.

Garrett Sussman: You’re not wrong. You’re not wrong. Yo, thank you, everyone, for chilling with us today. We’ve got to do more of these because right now there’s just so much to talk about. Everything is happening constantly.

Well, obviously, right after this, I’m going to throw this in my Mac Whisperer. I’m going to get the transcript. I’m going to have ChatGPT restructure it. We’re going to put the video right on the website so you can go read and watch.

Mike King: One of the Google people had said, “Now all content is infinitely remixable.” And I’m not just going to attribute that to him. Ross Simmonds has been saying that for the longest time. But with generative AI, there’s no reason that you can’t take this one piece of content we have and turn it into a variety of things. I want to see a cartoon. I want to see a dramatic film out of it. I want to see 52 images. There is no reason that we should continue to limit ourselves to the formats that we’ve always done.

Someone said to me, “That movie we made, Runtime, imagine if we had all these tools that we have now to make Runtime.” In fact, we should probably make a part two based on all this.

Garrett Sussman: I mean, that’s exactly why I was going to… I was having a little fun with Veo already. It’s awful. And yeah, it’s just minimal engineering. I forgot that I had to tell him about the long hair. This goes to the context. You have to be very specific with what you’re looking for. But this is the point.

That’s why I need to bring back Rankable, start doing the SEO Weekly. All the newsroom graphics that I wanted to do, I can do so much faster, so much easier. Y’all—

Mike King: You guys heard it here first. SEO Weekly is about to get a lot cooler.

Garrett Sussman: There you go. And yo, stop judging people for using AI content. I’m sorry, but get out of here with that. You’re better at content than AI. It’s a tool. You just make sure content is better. So own it.

All right. Thank y’all. We’ll catch you later. Garrett Sussman, Mike King. iPullRank. We out.