How AI rewrites search for publishers

(upbeat music) - Hello, hello, and welcome to another episode of the Digi Day Podcast, a show about the business of media and marketing. I'm Kamiko McCoy, senior marketing reporter here at Digi Day. - And I'm Tim Peterson, executive editor of video and audio at Digi Day Media. What's up, Kamiko? Welcome back. How was your week off? - It was great. I saw a several notifications about Google and I promptly turned my phone over waiting for this conversation. How are you? - Yeah, you definitely had a better week than Google 'cause the hits keep coming from Google. I mean, even this morning we were recording this Monday, September 18th, Google just got hit with a new lawsuit from Pomadec. So Google is just getting it from all sides, which is kind of what we talked about in the summer recap episode with Digi Day managing editor Sarah Gerty. So. - Heavy sigh from everybody. A collective heavy sigh. So later on in this episode, we're gonna have our senior media editor, Jessica Davies, and our senior media reporter, Sarah Guaglioni, on how AI is affecting publishers traffic, which is apt given, you know, we're talking about AI search and all of the insane through lines that are happening in that space. But first, like I mentioned at the top of this session, is this week's Juicy Scoops is dripped out in a G shaped hellscape. - Yeah, yeah, there's a lot going on. I mean, you know, the Cindy Rose era, WPP has officially started YouTube just, well, YouTube's part of Google, but YouTube just had its first NFL live stream outside of the Sunday ticket package that YouTube has. But the biggest, I mean, far and away the biggest story this past week, at least as far as our coverage is concerned, was we finally got the remedies in the Google search antitrust trial. So for anyone who hasn't been following or just hasn't cared to follow, year ago, Google got found as having a monopoly when it comes to search. And so then once that verdict came through, it was kind of a question of, okay, so what do we do about this? And because you had talked about on the show, there were a few main options. One that the U.S. Department of Justice was really angling for was Google being forced to divest Chrome. So that was one under the spectrum. Other end of the spectrum for the cynics out there was Google's just gonna get slapped with some fine and kind of continue as normal. That didn't exactly happen, but neither did Google being forced to divest Chrome. Instead, Google just has to share some search data, not all search data. So that's an important caveat here. And it also can't pay for to be the exclusive search engine on platforms like Apple Safari browser, but it can still pay to be a search engine built into these browsers or kind of embedded in these browsers and other platforms. And so there was kind of a collective face palm, I think from the industry last week. - Is because to me, in all of this, mind you, there obviously in everything that happens with Google, there are some that, you know, chicken little, the sky is falling type deal. But this felt like really in my interpretation, not like quite a slap on the wrist per se, but not much better than a slap on the wrist for a monopoly. And it seems like there were some ad industry execs who maybe felt the same based on run and shields reporting. - Yeah, senior ad tech reporter run and shields was all over the story last week and talk to a bunch of people who were just fairly pissed off about this, also disappointed about this. And then some of them were just like, yeah, this is kind of what you come to expect, which, you know, having covered tech regulation for what a decade and a half at this point, this is what you come to expect of just not a ton changes. And so one of the I think more disappointing parts of this for people in the industry is in two weeks, the remedies hearing in Google's other antitrust case, the ad tech one starts up. So like with the search antitrust case, Google was found as having operated an ad tech monopoly. And so now it's a question of, okay, what do we do about this? And similar to the divest Chrome conversation, it's do we make Google divest its ad tech business? It's, you know, formerly called double click. Or do we just make Google pay a fine or do we have Google sell off parts of it? Or is there like some sort of like in between remedies? The search one though was seen by industry executives, according to Ronan and this morning, I was talking to our executive editor of news, Sub Joseph who said the same thing. The search trial was seen as the big opportunity to change things up, to really penalize Google for having this monopoly in search. And so because the remedies were relatively weak, it's like really taking the wind out of the sails when it comes to any sort of expectations for the ad tech antitrust remedies. Yeah, which to me, there's the potential for that to kind of like set a precedent as we move forward. Now, also to your point at the top of this conversation, if Google avoided that ball in dodge, this game of dodgeball, right? There's still several others that are coming at it, including a lawsuit from Pupmatic and then a fine from the EU for ad tech violations. So not to say that they're out of the clear yet. The opinion just filed find Google 3.5 billion for ad tech violations late last week this morning. Like we mentioned, Pupmatic's filed a lawsuit against Google alleging an ad tech monopoly last month, OpenX filed a lawsuit against Google alleging auction rigging. So Google is just getting it from all sides. I mean, the $3.5 billion fine, that's a lot of money. It's not as much money to a company like Google that's worth multiple trillion dollars. And it's also the kind of thing of like, okay, yeah, you're finding Google for this money, but that isn't going to change the nature of its monopoly in the same way that the search antitrust remedies isn't gonna change the nature of like Google has a monopoly in search. And so it's all of these things are like, Google's still getting it, but Google's kind of just shaking off all these assaults. - Yeah, and that's the thing that there are several assaults happening at one time in Google and I don't know. Is this a too big to fail situation? - Isn't everything, I mean, ever since 2008, like I just feel like everything that the innovators dilemma doesn't feel like it exists so much anymore. It's just you gotta get bigger. - Geez, that's the innovators dilemma. - Well, we will obviously be keeping a close eye on the lawsuits, the several lawsuits and ad-tech violations coming from both across the pond and here, but all that to say bad day for proplexity, which was planning to put in a bit on Chrome. So the sun has been set on that at least. - Yeah, I mean, proplexity had a tough week in that respect, it had a tough week in that it's the head of its advertising business has left the company. And I believe it was ad age who reported on that and also mentioned in their story that like proplexity had only brought in like $20,000 in ad revenue in the first quarter that it started selling ads. - Yeah, and now they've also got some competition. - From Apple, yeah, so according to Bloomberg, Apple is planning an AI powered search engine. And so that would create competition for proplexity, for open AI, which has web search capabilities and chat GPT and potentially for Google too, although that's what we were just talking about. Google seems fine with competition 'cause there doesn't seem to be any true competition for Google, but this would be competition at least for the Google would be rivals like open ad and proplexity. - Yeah, kind of those, I guess you can call them the mid-tier players in this space where it'd be competition for them. And maybe that accounts for like smaller business and things like that and whatnot. Google still at the end of the day, obviously, to your point has like a massive, I won't even say a massive whole chokehold on this industry, but if there's competition-- - A monopoly, maybe. - If we want to put labels to things. But maybe a smaller player space here to that case. - Yeah, and it'll be interesting to see what actually comes of Apple doing an AI-powered search engine, 'cause Apple's made faints at getting into search. Like I think it was the middle of last decade that there was some search momentum in the same way that like it was, I think around the time that Apple launched Apple Maps. And so there were like a collection of things Apple was doing that seemed to be about creating competition for Google and lessening its reliance on Google and potentially lessening the payments it needed to be making to Google to be the default search engine. One thing that's interesting about Apple planning an AI-powered search engine is to what extent this is a defensive move in the idea that Apple could be getting less money from Google to be the exclusive search engine. Because then Ronin was the one who made this point. - If Google can't be the exclusive search engine, then it's not gonna be as valuable for Google to be a search engine. So why would it pay as much for less in return in terms of the exclusivity? And I mean, $20 billion a year. - How do you make that up? - I think it's what's been reported that Google's exactly. - How do you make that up? You make that up by making Siri, well, trying to make Siri worth somebody's time by infusing it with AI. I think internally, as Bloomberg reported, there is titled world knowledge answers is how they're calling this. So Wika, I don't know. - Which is right up there with like MS Now and Howdy in terms of like just the best names to come out in 2025. - Did I do think that all, I mean, it's an interesting play for Apple specifically because of Siri, you know, and maybe a potential to make good on what they were offering with their Bella Ramsey ad that got pushback about all the capabilities of Siri and whatnot. So to your point, it seems like this is probably something that's been in the wing or at least, you know, in the, maybe a glimmer in their product developer's eyes, just as the Google remedies were coming down the pipeline and what that would mean for, you know, their revenue streams. - Yeah, and still this question of like, can Apple deliver because it hasn't delivered on Apple intelligence? - Correct. - Also, is there still a part for Google in this because I think it was Bloomberg had also reported, Bloomberg or Reuters had reported over the summer that Apple was also talking with Google about making Gemini part of its, part of like Apple intelligence, part of its, you know, AI strategy. And then Meta is just, I mean, Meta has been rating everyone, but like Meta has taken a lot of AI executives away from Apple. And so, okay, Apple, cool. You're playing an AI powered search engine smart timing because you're about to announce the new iPhones and you know, roll out the new iOS and everyone's still waiting for Apple intelligence. So yes, you know, like kind of direct our attentions towards this AI powered search engine that may or may not be in the wings so that we forget about the fact that Apple just hasn't delivered on AI so far. - Yeah, I mean, or their products overall for that matter, but different story for a different day. Moving on to our very last topic here, which is also if you thought we were done talking about Google, guess what? We're not on our last juicy scoop. We're going back across the pond where there, there's some more beef with Google. Tim, tell me what's going on here. - Yep, so UK regulator, the competition and markets authorities already investigating Google's search business and that includes AI mode and AI overviews, but a group of publishers is now calling for the CMA to include Gemini. So Google's AI assistant, Google's version of chat, GBT, Claude as part of the investigation, which makes a fair amount of sense because as we've been talking about these AI chatbots do seem like the new search engines. And so it would make sense for Google to include Gemini as part of that. Obviously, Google does not want that to happen. It's trying to make a claim that it doesn't make sense to include Gemini as part of this, but this broader investigation has to do with the digital markets, competition and consumer act. And depending on what the investigation leads to, it could end up with the CMA requiring Google to have to pay publishers for content that appears on its platforms. And that would include search. That would also then include Gemini. So you can kind of see why publishers want Gemini as part of this because if Gemini does become the new Google search or at least as important as like Google Discover has become important as in terms of a referral traffic source for publishers, publishers want to make sure they're getting compensated for it. - I think what's interesting to me there is like in its search monopoly case, right? Google kind of positioned itself as far as I was reading the one time I did turn my phone over to actually look at this news while I was away, Google position itself was like, well, hey, the real competition is with AI. So what's all this business about us focusing on search? Let's talk about the future. And then here comes the UK publishers being like, yeah, indeed, let's actually talk about the future right now in the monopoly that's happening here with AI and search and also pay me up for all of this information. - Yeah, which is really savvy for them because I mean, it is, if not the future, a major part of the future as far as we can tell. So it's smart on them to say, all right, if you're investigating Google search business, you gotta investigate kind of the next generation of that search business, which is Gemini and which also tees up perfectly the conversation we have with Jess and Sarah G in the featured conversation this week talking all about, you know, how things like AI mode, AI reviews is affecting the traffic that publishers are getting from search engines like Googles. But then also what publishers need to be doing in terms of figuring out, okay, what's the strategy going forward? Like this stuff is happening so you can't, you know, bury your head in the sand. And I love Jess and Sarah and you as well, made a lot of great points in terms of what, how publishers need to be thinking about what's going on right now with respect to AI. - Absolutely. It was a real good conversation. So with no further ado, what's happened? (upbeat music) - Jess, Sarah, welcome to the podcast. - Thanks so much, great to be here. - Yeah, thanks for having us. - It was, it was like to have you 'cause Kamiko and I have been talking a lot about publishers in AI this summer for a good reason. But you both are the experts there we had you on recently to talk about the IB Tech Lab meeting with publishers. But since then it feels like there's been a bunch going on. You know, Jess, you recently reported on just how, to what extent publishers traffic is being, or seems to be affected by AI overuse. Is that, I guess is kind of the thing here. If there was no impact on publishers search traffic, there wouldn't be as much angst around, what is AI? What do AI overviews from Google AI mode from Google? How is that going to impact publishers? But it seems like just there are impacts to publishers being seen. - Yeah, completely. I mean, it's never a dull day as, as you can see, there's, there's changes happening just constantly. But it feels like, I mean, volatile referral traffic, it's kind of become the norm, hasn't it, for publishers. But the AI stuff, I think, definitely adds a new layer onto that. And yeah, it seems like clicks are drying up. There's no denying it. You know, the data backs it up. There's more and more data coming out all the time. And I think, I mean, I feel like for a while, Sarah, I feel like you, you would agree with this 'cause I know you've reported on this, but for a while there was, you know, a feeling that like the usage bases of the AI platforms like Open AI and Perplexity, they weren't really large enough, you know, to really worry publishers. And so that was always something that we were hearing. The, obviously, the main one that everyone had their eye on, that everyone had the fear over was Google, just because, you know, AI overviews, it's built into search. So the user base is there, it's established, it's massive. And that was always the thing, you know, that we were hearing constantly, but all of it together has obviously been a worry. And Google, I'm sure this will come as a big surprise, but Google's been a little bit opaque in how it decides, you know, what gets cited and what it's really meaning. There's no kind of like, I guess there's no Google analytics for AI usage in a way. So that's definitely been a challenge, but there's a lot more kind of third party data that's been coming out. We've reported a lot on it. And I guess the latest one that we had came from Digital Content Next. They did a look in May and June, they asked about 19 of their members to take a proper look. And I think it was like maybe four or five days or a week roughly after that Google post that Liz Reid wrote, Google's head of search, Liz Reid, she wrote quite an extensive post saying that a lot of this third party data that's being reported everywhere is not accurate. You know, they are actually sending higher quality clicks, but there wasn't really a lot of data in that that could really back that up. So I remember when I was talking to Jason Kint, he was putting together this report. He was saying that this was kind of like a ground truth, you know, for publishers to sort of set a line on what the real data was. So... - Is that true though? 'Cause that was one thing I wondered here is we had, Sarah, you were in the room for this, the digital publishing summit in Vale of this year during one of the town hall sessions, we talked a lot about referral traffic and kind of search traffic. And at that time, there was a core update Google had made to search like within a couple of weeks prior to DPS. And so there was a fair amount of angst in the room from the publishers on what's this mean for our traffic? What are the impacts everyone's seen? And I believe there was one publisher who made the point that like, if it's a single digit swing in either direction, that's kind of within the margin of error of, it could be from in that case, you know, the core update or it could just be kind of an organic attrition. And so in reading, you know, the stats from digital content next where they, they looked across, you know, eight weeks and they found that the median year over year decline in referred traffic from Google search was 10% drop overall, 7% for news publications. And that made me wonder how, because then the next, you know, paragraph in the piece is Jason Kintz saying, this is the ground truth. But I had in the back of my head this publisher saying, if it's a single digit decline, that's kind of the margin of error. So is it, does that make things, it feels like everything about this is super marky at the moment. - No, it's a good point. I mean, I feel like, you know, just to chime in real quick, we had a, you know, ask the editors event, you know, yesterday we had Ed Hyatt, head of SEO at the Wall Street Journal, you know, come and talk to us and dig day plus subscribers, you know, members. And he actually said to that, at least for them, there was a bit more volatility recently from the recent Google core update or actually, you know, not just the recent one, but the past couple of ones can be more, you know, influential in some ways than what they've seen at least with AI overviews. So I still do think it's a widespread problem that's being caused by AI overviews. I just think that maybe the degree to which they're really existential or not really depends on, I think the publisher, what kind of keywords they were targeting, how much evergreen content, you know, they were, was part of their sort of editorial strategy because, you know, if you were doing a lot of evergreen explainer type of, you know, stories, I think you're definitely more at risk of losing a lot of your traffic. Like someone else in our publisher virtual town hall last week was saying they lost 50% of their traffic since AI overview. So, you know, you hear numbers like that and it does kind of, you know, make your eyebrows shoot up into your hair. - That person, 'cause then we asked that person, okay, what drove that 50% drop? And they said it was a change in like their syndication partner. So it wasn't AI overviews that they attributed to. - This comes back to the tricky murkiness that you mentioned because it's the same here, you know, in Europe, like a lot of publishers are reporting like higher and higher click-through drops and sort of general referral drops. But there's, it's really tricky still to get that clarity and transparency on the data and the measurement. They're sort of, I mean, I think in May, you know, MailOnline spoke at a conference in Poland here saying how they were seeing like, you know, 50 plus percent drops, you know, so that was already back then. But as Sarah says, it really does depend on the publisher and there's just, there's just still a real gap there in terms of what they can actually see and what they can therefore then tie it to. And then that can help inform strategy around what they need to do with SEO, whether they need to kind of push harder on how they're being compensated, you know, because they don't really have a lot of, or as much insight as they'd like, right, into sort of how it's being used. - As we're having this conversation, that's kind of like, that's the voice in the back of my head this entire time, right? It's like, if there's so much uncertainty and murginess and opaqueness, like how you determine as a publisher how dire the situation is, right? And then to your point, just determine like, how to strategize around this? Because if you don't know like, what pillars that this is affecting, where the traffic is coming from, where it's not coming from, to be able to then, 'cause it costs money to come up with a strategy, you know? And publishers are already kind of being squeezed here. So that puts you in a tight spot to determine, you know, what the strategy looks like going forward. Has anybody kind of come up with a path here? - Yeah, there's sort of a way to piece these things together. Yeah, Jess and I have covered this a bit. It all seems a bit piecemeal, to be honest. It's a combination, I think, of publishers own, you know, search data. Again, there's no AI specific data in there, but just sort of seeing trends, you know, what's doing well, what's not. And then comparing those, you know, queries, keywords, to sort of third-party tools, like, you know, SEMrush and Profound, I think were a few that have been mentioned to me, like SEO tools that can help kind of show visibility in places like AI overviews is what it sounds like to me. So you can kind of match the data and kind of get a sense of like, okay, we're showing up a ton here, but our traffic is like suddenly, you know, falling off a cliff. I think it's, I'm kind of oversimplifying it, but it does seem like that's the way that some publishers are sort of putting, you know, painting a picture, right, Jess? Like, I don't know if you've heard-- - Definitely, I feel like, and I, this is sort of my interpretation of it, obviously, but I definitely feel, I sense a bit of a mood change, you know, from those kind of earlier months when everyone, you know, the conversation was very much around like, oh my God, you know, this is crazy, and yeah, we're still trying to, still trying to get more data on it, we're still trying to see the impact, and now it's a lot more around like, okay, I think everyone can kind of agree that the writing's a bit on the wall, even if it's not right there now for them in a kind of, you know, this is gonna go crazy tomorrow kind of thing, but it's definitely almost like a kind of, oh, here's the latest monkey wrench again. So proud of knowing that term, by the way, we have a different term over here, I had a real-- - What's your term, spanner? - Yes, ah, you know it, I'm so impressed, yeah, throw a spanner in the works, yeah, absolutely, but I'm sure I got tied up in like, it got very tangled talking to-- - You've been American, I've-- - Absolutely, I have, I have, even sometimes here, some of my friends are like, what? I'm like, yeah, but I love your term more, monkey wrench is great, but anyway, yeah, so it's kind of like that feeling of, okay, you know, we're used to having to kind of, you know, adapt on the fly a little, so a lot of, even, you know, Ed Wyatt, you know, as Sarah already kind of mentioned from Wall Street Journal, he we spoke to yesterday, he was very kind of optimistic, really, about, you know, there being ways to just make sure that they're focusing on, you know, the quality journalism, not commoditized content, and that's something that everyone, I mean, that's obviously been the, you know, the aim for most publishers for a long time, but now it just feels like it's a, okay, no more SEO gaming, like let's focus on, you know, direct connections, and, you know, that kind of thing, and visibility, brand visibility seems to be something that people talk a lot about, being an important part, that it's not just about SEO or adapting SEO for AI, you know, GEO, whatever, if you're using all the various acronyms, but it's about the brand visibility and the technical capability within that. So there are definitely sort of some good strategies coming out, and I feel like there's a bit of a mood shift. - Yeah, I think that's a really good point, 'cause it almost feels like some publishers have accepted the fact that they're never going to see the same traffic that they've seen in the past, ever since AI overviews has really, yeah, has really hit and, you know, chat to BT and perplexity, sort of the threats that they pose, and I agree, Jess, it almost seems like, okay, now what? You know, like, we've tried everything, and it's just not gonna come back the way that it once did, maybe because of AI, maybe not, we think it's because of AI, but either way, we have to figure out what's next. - You said that's so cheery, and that is such a dire, (laughing) sorry, is the end of our traffic as we know it. - Well, not the end, not the end. I do wanna clarify, I don't wanna say it's the end, 'cause, you know, like we're saying, like some publishers are still seeing traffic coming in through search, 100%, you know, good traffic coming in through search, but not these huge, you know, it's not growing the way that it once was, and people that have seen large ships, they're like, we've tried all these different things, and it's just not gonna come back the way it once did, so there are different tactics for how to figure out, you know, what's next, looking to different platforms, like Jess said, trying to drive really direct relationships with publishers, trying to squeeze more revenue out, you know, per reader. - Well, on my end of the house, for like marketers, right? Because they too are thinking about this from like their own brand product pages and whatnot, right? And there's a sense of like, I've heard the word authority more times than I can like shake a stick at, right? (laughing) We need to be an authority on all of these pages across our blogs, across our socials, so as these pages are getting like crawled by these AI bots, there's some type of consistency and we show up. But the thing that was interesting is like, putting more strategy around things like blogs and like influencers and whatnot to capture a bigger share of the market for AI to crawl. So the question is like, what does that look like from the publisher side of the house? Whether that be like a Reddit or blogs themselves, like that type of thing to kind of capture some other referral traffic. - Yeah, totally. I think publishers are doing that too. They're increasingly creating networks of creators that they're tapping into and trying to work with to create content that, yeah, like you said, isn't reliant on these more traditional, I think search channels. They're looking to places like Reddit for community engagement, but also referral traffic, finding the right sort of subreddits where people are really interested in the content that they produce. And it's actually been a growing referral traffic channel for some publishers. And even Facebook, which is kind of wild because, you know, - Deja vu. - We can have a whole separate episode about Facebook, to be honest, but, you know, - This is PTSD. - I would say Facebook, you know, it's been a very challenging relationship with publishers in the last couple of years, you know, with all the things that they did around deprioritizing news and taking money away from publishers that they were previously giving them, I think a lot of publishers that kind of lost hope. But then all of a sudden, beginning of this year, they started to see sort of a resurgence of referral traffic there. - Yeah, we've seen that too at Digida 'cause I ever see our social team. And so we've seen Facebook traffic bump up this year, but the conversations we're having are the same ones that all the other publishers are having is like, I don't know how much we can trust this, so I don't know how much we can really try to take advantage of this, only to be left holding the bag in a few months when Facebook changes its mind. - Totally, yeah, I feel like publishers that told me that they saw this uptick in Facebook referral traffic, they almost wanted to like whisper it, you know, like, yeah, yeah, we've also seen a ton, but. - Yeah, maybe it's accidental, don't you? - Yes. - No, no one. - Exactly. - Yeah, that's so funny. It's a funny thing in terms of, I wonder whether there'll be more there from Facebook in terms of, you know, we've had a few whisperings just about how, you know, I mean, it's very much like just whisperings, there's nothing more concrete to it than that, but just, and I'm sure you maybe kind of touched on it in the previous podcast, I, so I haven't had a chance to listen to that, but I will when you talked about the IAB's workshop, but there was some talk around Facebook potentially sort of just having a, you know, being open to more conversations with publishers, you know, when it comes to things like AI licensing and everything, nothing confirmed, but it'll be interesting to see, to watch that space and see if there's any kind of change. 'Cause one thing that sort of, and not to kind of go too, too often a tangent here, but one thing that I thought was really interesting in a conversation I had with Anthony Katzer, who's the CEO of the IAB Tech Lab, we were just talking about what was funny in terms of who turned up to that meeting and obviously the AI platforms weren't there, but Meta and Google were, and it was a kind of thing of, well, yes, okay, there's all this tension always around the relationships with Meta and Google over the years and continues to be so with AI, but at least they come, you know, at least they do engage and they want to engage. And it's just that, I just thought that was kind of interesting. They really are these like new players that are not really familiar with the same kind of, I guess, the frenemy dynamics, you know, that have existed for a long time. And so they just don't engage, you know, unless there's obviously like a specific licensing deal that they want, but yeah, I just thought that that was a kind of interesting thing to kind of look at. - Yeah, I remember one of the attendees told us, you know, to that event that, you know, those big tech companies were there because they know that, you know, they should show up and at least try to be part of this conversation and try to work with publishers on something versus startups that, you know, maybe still think that they don't need us type of thing. Maybe that's a bit of an optimistic perspective, but I thought it was an interesting one that they shared that, you know, because of the history, you know, these larger sort of more institutional platforms knew that they had to be in the room. - Although it makes me wonder if like what we're seeing across all of this between which AI companies are in the room for these IAB tech lab meetings or feel like it's necessary for them to be meeting with the publishers through IAB tech lab, as opposed to doing individual deals with big publishers, as well as, you know, just the impacts when it comes to things like AI overviews, AI mode, is if what we're seeing is a calling of publishers, you know, like one thing in the digital content next study that Jesse reported on is how it was a negative 7% drop for news brands and negative 14% drop for non-news brands. And that feels indicative of news is more of a utility, there's more use to that information. And also, there's a focus group of one, I'm not trusting Google Gemini to translate the news for me because of the hallucination issue. But if I'm looking for a recipe for something and my option is to take whatever AI overviews regurgitates or having to go to one of those freaking recipe pages where it's like 600 words of the person talking about their great aunt and the relationship that they had and what was like growing up with them before I can get to the actual recipe and I have to go through 12 in-stream ads in the process, I will take the AI overview every day. - That is a solid point. - Well, I guess that's going to be down to, a lot of these food publishers, big magazine publishers, they are innovating a lot with their own LLMs and doing actually some kind of, I think quite interesting stuff in terms of how they resurface their own stuff and just make the experience better so that they're actually providing that in-depth stuff that isn't necessarily without all the kind of, the sort of the ancillary stuff around the side. But yeah, I would say that that's on them to kind of just adapt to that. And I think a lot of them are from conversations that I've had with a few of those types of publishers. I don't know if you've come across that Sarah, but... - Yeah, I think it is going to be a bit of a survival of the fittest, which is sad, but really looks like the future. I mean, Dot-Dash Meredith now called People Inc. The CEO even said this himself that the core, their focus is on the core titles, and their big lifestyle, food, travel, publisher, their focus is on their core titles. And the other ones that are being hit by all the traffic declines that we're talking about here, they're just not going to be putting any more resources into those titles. And what does that mean? That means, are they just going to fizzle out and die? I mean, it kind of sounds like that. And so I think, unfortunately, it does seem... The good news is that it does seem like news publishers are kind of insulated from some of what we're talking about. The bad news is those lifestyle recipe, food blogs, the recipe that you're reading comes from those sites, right? So if they're no longer around, what does that mean in terms of, yeah, content in those spaces? - Although I wonder if the future for them, the future for kind of publishers we're at large, to an extent, is you use the term survival of the fittest. If everyone just has to get fitter in a way in order to survive, like those recipe publishers, they still have a value. They, I just don't need the 600 words about your grade on. Like, that's nice for you, that is not helping me 'cause I got dinner to make, now I'm hungry. But so at the same time, like AI overviews, Gemini, well, maybe it's learning how to make, I don't know, pan-fried salmon or what have you, but that information's gonna have to come from somewhere. Recipe sites are a great source for that. And this seems to dovetail with like the incentives that are starting to align that way, where these AI deals initially started off as just like broad content licensing deals. And now you both have been reporting how there's this, seems to be this shift towards usage-based pricing. And I wonder if that opens an opportunity for publishers, especially for the non-news publishers that are more, seem to be more directly being affected of if they can be creating more useful content and that's their path for survival. Just what do you think? 'Cause you've been all over this topic. - Yeah, I definitely think that's true. I mean, and I just saw something earlier today, actually a report come out about how much sort of AI slop has been created and recipe sites, you know, there are sort of tenor penny to use very British expression. But like, I spoke with someone, I think just a week or so ago, who was looking into, digging into some data on this. And there was just an insane amount of sort of AI-generated sort of nonsense sites, and they're all food sites. They just have the same pictures and images going. And this report today just came up saying, well, this could actually be a good thing, you know, for quality, you know, quality journals, you know, there's usual trying to find those silver linings. But it made me think, well, yeah, actually, this could flip around like if it comes to a head like that. And there's just so much sort of, you know, garbage on there. Then this could be a real place where, you know, proper, proper titles doing it right will kind of shine through. But yeah, in terms of the licensing, I mean, it's been really interesting seeing that because when I first started asking people about it, I was kind of saying, you know, in the early days of these licensing days, and in the early days, I kept saying like in the early, you know, and then I look back at when these actually, when these actually were, I was like, oh my God, this is just last year. And like, this is crazy how fast it all, you know, the time is compressed, that's how it feels. So I had to start changing that and be like, so the early deals, you know, that were kind of actually just recent. But yeah, from what I've gathered from talking to people who are, you know, obviously closer to it than I am, specifically, it's just that, yeah, the sophistication of these deals is changing now because everyone's had time to learn, you know, about what the hell is going on, basically. And, you know, they've time to kind of refine what they want and think about publishers on their side as well as the AI platforms on their side, I guess, sort of thinking about what will work better and what's a better kind of more sustainable compensation model. And now that everything is based on, you know, that real-time information retrieval, it does seem to open up a lot more than just those early days of, oh my, you know, the fear that was around then and that desire to just like, well, we need something to kind of make up for what you've already done and what you're continuing to do. Whereas now it's like, okay, well, what's the longevity here? We all want, you know, the right economics here in terms of a sustainable web. So let's make it work. And it definitely feels like there are a lot more sort of just in pockets. I hear it, you know, just things going on in different places where I'm like, oh, that's an interesting thing. That's going on there and, you know, different types of things. And obviously there's an umbrella term, you know, paper usage, but there's a few iterations of that. So yeah, it's going to be interesting. I'd love to be a fly on the wall, but these deals, but... - And Sarah, you just reported on Perplexity introducing this new program, Comet Plus, which has usage-based pricing baked in. If people are paying for this to use the Comet Assistant and Perplexity's Comet browser, if a publisher's content gets cited in the results or a person clicks through to visit the publisher's site, that publisher would get some cut of the $5 a month a person's paying for Comet Plus. And so that is based on usage of the content. Is this Perplexity spearheading a new model, or is this Perplexity, which has, is kind of in the tier below an open AI, an anthropic, a Google, a meta... Meta, what exactly tiered they're in when it comes to the AI in the large language models, but is this Perplexity kind of scrambling to cozy up to publishers? - Yeah, it's a great question. And I think it really breaks down to this idea that both of you guys kind of laid out, which is those early deals were really, and an exec told me this, were a bridge to get us to where we're coming to now. It was like, we need money, AI companies are scraping our content, we need to be compensated for that, give us money, and they did. But it's not a sustainable long-term revenue share model or really, you know, model for internet usage in the age of AI at all. And so I think what's happening here is Perplexity, you know, they launched an ad revenue share program last year, it doesn't sound like it's been very lucrative at all for publishers yet. And I think that's because their ad business is just not really taking off. I don't think there have been a lot of resources that have been put into it personally. And so they had to figure out another way to compensate publishers for, you know, crawling content, especially as you're facing a lot of legal pressure. And this kind of model, this pay-per-use model seems to be the new trend in this space. You know, it's not just Perplexity that's doing this. I think it's one of the big news of the week is that it was Perplexity's model that really came about. But the Ivy Tech Lab has been talking about this too, you know, with Cloudflare, with Tollbit, you know, these other tech companies about setting up a model where publishers get paid when specific content is taken by an AI platform to answer a query, right? It's not a big licensing deal. It's really just per usage. Tollbit actually already kind of offers this. It's just, I think, not really taken off yet. And then, you know, they're part of what Ivy Tech Lab is trying to set up too. So to me, it really sounds like it's the start of this shift, a real start of the shift as these different companies are setting this up. I think what's interesting is that it's all kind of happening at the same time. So I'm like, are there just gonna be multiple paid per use programs for publishers to make money through AI companies? And does that get really complicated? Or is that a good thing? Because then publishers, you know, have the opportunity to make money from many different places based on where their content is coming up. - So it creates competition for better terms. - Very good point. Yeah, very good point. And I think that was the issue with the Perplexity offer. Was that, you know, the model itself was very clear, which the people that I talked to really appreciated because it was like, here's this large pool of $42 million that you guys can have access to. And this is how you can get in, you know, from traffic through Comet. The real question that execs had to me was like, okay, but like, how are we getting paid? Like, is human traffic, you know, weighed more than bot traffic? Like how often, you know, are we even gonna get traffic from the browser? Like, you know, I think those kinds of terms were not clear to them. They hadn't met with Perplexity yet to talk about them. And I think that's all kind of being worked out. And I'm sure, you know, that will be a big part of the Ivy Tech Lab proposal as well as like how to really set the rates for these kinds of things. And like both of you said, I think the competition part of it will be a big key part of that. - Right, Jess? I mean, what do you think? - No, I do. I mean, essentially the, so I think there are a couple of things like, so the Ivy Tech Lab favored one, I think is the paper query, which is kind of like paper usage, you know, whatever. It's basically you're getting paid every time a user says, search is something and your content gets pulled into that summary or citation. And the other one is the paper call. And although a few people have sort of said they prefer the paper query, there could be potentially, you know, some at least short-term revenue that could come from the paper call because the AI callers, they do actually go repeatedly to sites, not just kind of like once a day kind of thing. They're going like, you know, potentially hundreds of times, you know, a day when they're retrieving information for every new type of query that's put in, they're going back to certain sites and things. So there could be some, you know, some legs to that. But yeah, it's going to be interesting to see kind of how it all shakes out. - I think typically I leave these conversations, especially about AI and Publishers Pilling, very defeated. It is very, it is very, it's an inception-like thing to be talking about Publishers and AI as a Publisher going through AI. But I really appreciate you guys because, you know, having this conversation, I feel a little bit more hopeful. There are some through-lines here in terms of like, you know, shifting compensation models and, you know, although there's a lot of market is to be dealt with, you know, there's Publishers have a seat at the table and, you know, that's all we can ask for right now. But Sarah, Jess, thank you guys so much for joining us on the pod to kind of let us pick your brains. We love to have you. - Thank you. - Thank you so much, it was fun. - Well, that brings us to the end of this episode of the Digi Day Podcast. Thank you to everyone for listening. And please don't forget to share this episode with someone who you think would enjoy it. You can even rate us and leave us a comment on Apple Podcast. We'll be back next week with another episode of the Digi Day Podcast. Thank you so much for joining us.