How AI Ads in ChatGPT Are Reshaping Advertising

This episode of BIA’s Leading Local Insights Podcast explores OpenAI’s move into advertising within ChatGPT and what it could mean for brands, agencies, and media sellers. Listen as BIA’s Celine Matthiessen and media analyst Mike Boland unpack emerging AI ad formats, evolving measurement challenges, and the expanding role AI assistants could play in search, discovery, and consumer decision-making.

Celine and Mike also examine how quickly the AI advertising market is taking shape, with ChatGPT ad pricing already beginning to move closer to premium Google search inventory as advertiser interest and targeting capabilities expand. They explore the rise of “zero-click” AI search behavior, new attribution challenges for advertisers, and why media sellers should position themselves as trusted AI education resources for clients

Transcript

Welcome And The Big Shift

Celine Matthiessen 0:34

Hello, and welcome to BIA Advisory Services Leading and Local Insights Podcast. My name is Celine Matthiessen, and I'm the VP of Insights and Analysis here at BIA. Today we're unpacking a major move that could change how ad world thinks about influence and discovery. Open AI is way into advertising with Chat GBT. That's a lot to unpack here to make sense of what's happening and what it means for all sides of the industry. I've tapped my colleague and expert analyst Mike Boland. Hi, Mike. How are you?

Mike Boland 1:03

Hi, Celine. Good to be with you.

Why ChatGPT Ads Matter

Celine Matthiessen 1:06

Good. I'm so glad you had time today to help us unpack this. So let's start at the 30,000 feet. OpenAI, the company behind ChatGBT, which everybody pretty much uses now, is now introducing advertising directly into the product. For those who haven't followed every headline, what's the big story here?

Three New ChatGPT Ad Formats

Mike Boland 1:26

Yeah, so so you said it. OpenAI is is jumping into ad monetization. Um and they've made the sort of surprising and unsurprising move to enter this, you know, why world of ad monetization that we know and love. Um and I say unsurprising because it's it's sort of a logical monetization path for something that runs parallel to search in a lot of ways, in that it's like a high-intent medium. Um, but it was surprising because Sam Altman in the past has been quite sort of outspoken in his public comments um about sort of the evils of ad support when it comes to you know maintaining trust and the objectivity in in like a search engine or something in this case that's sort of sort of like a search engine. Um and those comments in the past were were a not so subtle jab at Google. Um, but now it's you know jumping into that world with both feet. And it was announced at the beginning of the year, and it's been sort of slowly rolling out since, and we've been observing what that looks like in the ad units and how they're targeting and stuff like that, which is all still evolving. Um, but that's sort of that's the background for now.

Celine Matthiessen 2:37

So, yeah, so for speaking of ad units, for people in ad sales who haven't seen these new units up close, what exactly is OpenAI offering inside Chat GPT? And are we talking about the same old display and banner ads, or is it something new? Um can you give us a little insight?

Mike Boland 2:53

Yeah, it's it's it's totally it's a little bit of both. Let's walk through sort of what it would look like from the user perspective. So, first of all, it's sort of modeled after search in that these are ad units that are targeted on a combination of contextual triggers, so the content itself, right? And then behavioral. So that's the user and their search history, and we can sort of get into more of those targeting mechanisms. And these ads will be shown to essentially everyone except for those that are on Chat GPT's like highest tier. I think it's the $200 or the pro tier. Um, and what it looks like is there's essentially three sort of formats so far. And I think these will evolve, but at the outset, there are three. So the first one is like what they call a sponsored recommendation box. And that's the one that's sort of most like a paid search ad unit, where it's sort of like an uh an ad that appears below the organic response and it's sort of called out in like a tinted box, so you know it's sponsored. That's sort of the thread that you'll see through all of these, is that they're big on making sure like it's known this, this is this is an ad. Um, the second one is a little bit more sort of infused with the content. Like so that first one sits below. The the second one is sort of contextual links. So, so Celine, you know, and everyone else listening knows that when you do a search on Chat GPT or a query on that that or any other engine, um, they're good at well, I don't know if maybe that's an overstatement to say they're good at it. What I'll say is what they do is they do their citations, right? The citations, which is like, and I think that's a trust thing. They want you to know that, like, here's our source, right? We're getting this from somewhere reputable. So they they they list the citations, right? So this, these, the second ad unit is sort of a contextual link that's sort of infused within those citations. Um, and again, it's labeled that's sponsored, so you're gonna see all the citations in one of them or however many will be labeled. And then the third one is sort of purpose built for shopping-related queries, and that's where they're testing this sort of like image carousel or product carousel. Um, and and it's sort of a scrolling horizontal scroll. So without leaving the chat, you can see sort of, you know, it's a Google, Google um shopping does something like that.

Celine Matthiessen 5:03

Yeah, Google shopping has that right. Yeah.

Mike Boland 5:05

Yeah. So it's it's like that.

Celine Matthiessen 5:08

Interesting.

Sponsored Citations And Trust Questions

Mike Boland 5:09

So those are the three so far.

Celine Matthiessen 5:10

Wow. Um, so it's really not just a banner ad anymore, right? I mean, yeah, it's just the what really creep what seemed to me is the second format. Um, so like if let's I'll give an example, like if you Google like how do I get rid of fleas in my house, and a citation comes up with all the examples, but also then an ad for cleaning service or for you know a local pest ridden's company, right? That event eventually could go to that level besides the ad carousel. So what kind of like I mean, the citations are really just you know, you know, can be any kind of brand, right? At this point, depending on the queries. Okay. That's yeah, to me that's really interesting because we tr we think when we do chat GPT searches and you see the citations, it's giving you like this is our trusted source we're pulling from. And so there's there's a trust issue there that they're playing on.

Mike Boland 6:04

Well that will that erode the trust? Like is that getting too close to home? Like are they getting a little greedy in that in that second one? That's a good good question.

Celine Matthiessen 6:11

Speaking of go ahead. Yeah, no, because we're not finished.

Mike Boland 6:16

Well, I was gonna say, I hope that that flea search isn't isn't one that you actually did. I hope that was a hypothetical.

Revenue Pressure And IPO Rumors

Celine Matthiessen 6:22

That's for another day. Um, anyways.

Mike Boland 6:25

Okay, good.

Celine Matthiessen 6:26

No, you're good. Anyways, um, so yeah, speaking of greedy, why is open IA making this move now and what why does this matter to agencies and brands? What do they need to take? What's their takeaway from this? So for the products.

Mike Boland 6:37

Yeah. So I think no, yeah, that's good. Like we we talked about the what. I think it's good to step back and look at the why. Um and I think when we're looking at these things, whenever you look at the why, you get like sort of their motivations, which help us help us sort of triangulate where it's going next. But anyway, I think if we take a step back, like why did they do this, particularly after my earlier comment about like Sam Altman in the past had been so against this, like changed his mind. Um, and I think it's it's um there there are a few things there. So essentially, um this opportunity had had enough like gravitational pull to overcome those past objections, make him reverse course. And I think the first one is that ChatGPT has actually decelerated in the past year in terms of its growth rate for paid subscribers. So that's its Go Plan, the Pro Plan, and the Plus Plan. I think that's the wrong order, but there's basically three paid tiers and then an even bigger one for enterprise. But anyway, the people that pay for Chat GPT, that's the point, has sort of slowed. And I think that that really has come about because they're no longer the only game in town. It's like when Netflix used to just have like all the streaming market share, and now it's like there are so many. So like they're losing ground to anthropic, especially in the enterprise. I mean, Claude is running away with enterprise. Um, and and you add all of this up, and and they had to sort of diversify their revenues and accelerate this move towards advertising, which frankly was probably always on its roadmap, um, and and do it earlier. And I think another key point there is the timing is key because I'm sure you've heard all this. There, there's like a lot of rumors on its um a potential IPO on the horizon. So they really need to embolden their sort of revenue narrative, diversify a little bit more, counteract some of those market share declines, um, and have just a stronger story to take on the road. So I think that's that's sort of the why of it all. Um, but I forget the second part of your question.

Celine Matthiessen 8:30

Well, I mean, what's the early reaction from like the agencies and brands in the broader advertising community? So the why? Like what's the reaction looking like?

Agency Backlash On Pricing Terms

Mike Boland 8:39

Yes, so not good, I'll say. So um let me let me explain what uh what I mean by that. So um there the way they set it up was was a little weird. And I think it's sort of um, there was a lot of backlash to the terms of these these ad programs. And I think maybe in their defense, they did that because they knew it was a trial program and they wanted to keep it limited, like they had some big blue chip brands and they're sort of beta. Um, but anyway, there was there was a $200,000 minimum campaign spend to participate. So that, you know, sort of, you know, cut out large portions of the the buy side. And then the second one is that they were charging a $60 CPM, uh, which is, as you know, a lot in like in like ad uh online ad terms. Um, and then third, which relates to number two, um, the ads are sold only on an impression basis, as opposed to cost per click or anything sort of performance based. And I think the biggest objective among those is probably the $200, like, you know, $200,000 um like entry fee. But but I think that a lot of people, the one that's gotten the most attention is that last one, which is that, you know, the the online ad world has developed standards, as you know, Celine, and and and expectations on the buy side for really strong uh return on ad spend analysis and ad attribution. And I think it it just put this program um at a disadvantage right out of the gate by only offering impression-based metrics. Um, and it just doesn't live up to those standards. So the reaction to answer your question has been initially mostly negative, but I think I see some some ways out of that for them. And we're already seeing signs of that, of how they're improving that and and really starting to excite some other pieces of the ad ecosystem, whether it be brand advertisers or agencies or just the rest of the sort of ad tech world.

Fixes Coming: CPC Pixels Ad Partners

Celine Matthiessen 10:32

Yeah, so I mean that brings up so the onus for the agencies and for the brands is really um they have to be strategic on the um on the creative and the analytics side when they work with them. Right. So, I mean, how are they measuring success then? Is are they just relying on the impressions? What have you heard? Or what do you what do you recommend to a brand or agency that you know wants to work with Chat GPT?

Mike Boland 10:57

Yeah, so totally they're gonna have to get creative to sort of work their way around that shortcoming. Um, so I think a few things there. One is we're we're hearing some some signs uh that the that like the the campaign entry fee has sort of dropped. Um the Digiday, I I just read the other day from Digiday. It's they they they've they've heard from their sources in the ad world that it's dropping to 50k instead of 200k. So that's positive direction. And then the that infamous 660 CPM has reportedly dropped to like 25 to 45 range, which is still, you know, a lot. But anyway, that's dropped. Um, and then um more importantly, the way Digidate put this was that tracking pixels or CPC campaigns are now sort of on the table. So they're they're looking at it. So this is all moving so fast. So like within like a month, they've already sort of at least addressed those three major concerns. But I think another key evolution on top of those is underway, which is ecosystem development. And that gets to your question. So these were first available on just like a self-serve basis for like, you know, like I said, sort of a cabal of limited big brand partners. Um, but now OpenAI is rumored to be working with Critio and Smartify and StackAdapt and a few others for these like ad manager programs that will sort of do a better job holding advertisers' hands. And beyond the actual like first-party direct partnership with some of these players to build these systems, um, there I think we're gonna see a lot of sort of third-party ad tech sort of stuff come out of the woodwork. Like one I read about the other day is from AdBridge, and they're offering to convert your Google AdWords campaigns into chat GPT campaigns, which sort of takes away a big headache in that sort of like that learning curve. So, what they do is they sort of translate all of the strategies within your AdWords campaigns, your keywords, your target audiences, all that into Chat GPT campaigns. So that sort of lowers the headaches a little bit. So I think that we will continue to see agencies, ad tech players, others sort of like rush to this land grab to build Chat GPT ad stacks. Um, even though there are some of those, you know, hurdles that we just talked about at the beginning. I think this they'll figure those out, it will become popular. And I think that the ad, the whole ad ecosystem is going to be opportunistic in running into support it as like middle layers.

Celine Matthiessen 13:30

Yeah, it seems to just like anything else when they add advertising platforms take a tech move. We have a whole middle layer ecosystem or multiple ecosystems, right? So are there any risks? What are the risks? Like, um, does this a real trust in AI assistance or will this backfire for brands? I mean, you know, there's also uh, you know, challenges with uh making sure that the uh the ads don't show up by queries that are not something that the brand aligns with.

Mike Boland 14:00

Yeah. I think that, yeah, those like brand safety sort of mechanisms and and all that is definitely gonna have to be sort of worked out by both Chat GPT and that developing ecosystem we just talked about. And I think that, yeah, risk is a good way to put it. Like, I guess the risk we can segment that by like risk for whom? So for open AI, um, you brought up a good one earlier in that does this erode the trust in the objectivity, the underlying integrity and objectivity of the engine. And I think that, you know, absolutely, um, you know, ads erode trust in any sort of platform when they're added. And I think that when you're talking about ChatGPT or or just maybe AI in general, there are already a lot of trust issues because it's new, there's hallucinations, there's all this other stuff. It's also like on top of all of that, it's sort of it has some, I guess I'll call it political detriments in the fact that the technology itself is leading to a lot of job losses. So I think that like, you know, AI has a trust problem and an image problem right now. And like while it's grappling with all those issues, now we're adding in advertising to potentially uh like push push the mistrust even further. So I think that's something we're gonna have to see play out to see like how people receive it. But it is certainly risky for the brands themselves and and the that sort of middle layer we just talked about. Sure, I guess there's um because it's new and experimental, and you know, we you as a brand, you you have sort of trust in these established places where your ads are showing up and know that there aren't going to be brand safety issues. This is sort of a you know, wild west that they would be entering and have to experiment and deal with potentially some of the growing pains of perhaps some of those like unfortunate ad placements that that can can happen um in those types of processes. So so yeah, it'll be risky and experiment ex experimentation for for yeah, experimental for everyone.

Celine Matthiessen 16:06

For everyone, yeah. I mean, the users too, and users are accepting of these sponsored answers, yes. I mean, I know I I I've seen them and I've been like, okay, you know, I mean, all right. It must be if they're if they're dropping their, you know, their their budget minimum, their budget floor and their CPMs, so they seem to be seeing more opportunity at the lower end, right?

Mike Boland 16:27

So yeah, that would be a good translation of that move, I think.

Celine Matthiessen 16:31

Yeah, because I feel like those CPMs are right in a line with what what you'd pay for Google for a highly targeted real high value. High value, right? And so so they that uh seems to be I haven't I haven't had a chance to see on the advertiser and what they're what what campaigns are happening that are proving those numbers from for them. But I'm assuming that's what's happening, right?

Measuring Value In Zero Click AI

Mike Boland 16:58

So yeah, and and the that brings up another point too, that the analytics are still gonna be a challenge, right? Like that that was that one of those challenges I mentioned, which is that um they're right, they're doing CPM now. And I think like sort of reading between the lines, one of the reasons I think that they chose that is because that they believe that performance-based metrics may actually undersell the value or the impact of these ads. Um, so they've gone for that impression-based approach. I think that um that that comes back to something that is also like a challenge here, which is that you know that for the past like 10 years, the term zero-click search has always come up, come up in like came up in the Google world, where Google increasingly will provide either an answer through its knowledge panel or just some sort of call to action, like a phone number or transactional link, thus preventing you from clicking through to the advertiser's domain. And that's been like the the sort of world of publishers has been lamenting about that for a long time, that zero-click search. Now, of course, that's on steroids in the AI world because things like AI overviews or ChatGPT answers are sort of answering everything right within that SERP, I guess, or whatever the we're calling the equivalent of an AI SERP. Um so so because there's that lack of a click, or at least fewer clicks, and OpenAI knows it, I think it's trying to grapple with like how do we do the analytics or how do we how is it bought and sold and measured when there may not be a click? Like we if if we sell it on clicks, it may actually undersell the value because there are indeed impressions from these answers that don't result in a click. So I think that's something that both OpenAI and the other players that might integrate advertising, but also that like sort of ecosystem of supporting players are gonna have to maybe even like figure out a new metric, like an AI native metric that can capture the actual impact when there are indeed fewer clicks happening. Um maybe it's some sort of like probabilistic or deterministic way to connect those dots. And and as you know, Celine, the ad world has really gotten good at doing that over the past few decades when thrown some of these challenges, like for example, an increasingly convoluted purchase path, right? That weaves in and out of, you know, different devices and media, like the CTV world has gotten really good at doing that, as has the the direct out-of-home world, in terms of connecting those impressions with something that then happens in like the real world, like three days later, which is like was like an impossible feat that they were given. And they've gotten really creative about creative about sort of creating those models that come up with probabilistic, you know, um attribution. So it might be something like not that, but something that applies some of that same creative thinking and how they get around this dilemma of like zero-click search as applied to AI.

Celine Matthiessen 19:48

Yeah, I think the again with the convoluted path to purchase that consumers have these days, it's definitely going to be how do we how do we look back, maybe look back metrics or look forward metrics, right? Totally. That's that could be a you know, bringing up the you know, the opportunity for advertisers, um, you know, CTV, one of the things the measurement too is uh that's starting to emerge is are the attention metrics. So for sure. That I mean, because I love how you call you what did you call the chat TPT response where it's not a click through. I call it like a briefing. That's what that's my term for it. I'm like, it's my briefing, and then I can proceed, right? And maybe it's that proceed behavior because it always prompts you what would your what's your next move? And so they're measuring something on that. Um, you know, not maybe that's their it's not really a click-through, right? It's so maybe it's it's time interaction time too. Who knows, right? Totally.

Memory Driven Targeting Gets Personal

Mike Boland 20:45

Yeah, dwell time, yeah. Dwell time and could could be could be in there. Like they're gonna have to come up with some like Frankenstein of a sort of ad metric that like that brings together all these like known things that like that make sense here together. Um and and and it's interesting too. This this gets to a few things you mentioned. The sort of the way that these things are served and how how uh effective they are in their in their ad placement, and then also the risk and and trust thing, which is that one thing that chat GPT has been been really known for is having a memory. So, and this this freaks some people out and really like appeases and appeals to other people, which is that you you may have experienced this if you're signed in, um, it and it remembers your history. And and Google's always done this, but ChatGPT is doing it in this very granular ways. It will give answers within the context of the profile it knows about you and your past questions. So, for example, um, if you're looking for a place to eat or something to buy, it'll remember things like, you know, you own a dog or you have two kids or you have a nut allergy or things like that. And it'll answer within that context. And I think that that is really going to help it in terms of granular targeting and performance on the ad side. And maybe that will be part of the calculus for what we were just talking about in how these things are measured in terms of their effective um, you know, placement and and performance based on the fact that it's like, you know, really becoming this, you know, assistant for people um as they ask questions about their life.

How Sellers Can Lead With Education

Celine Matthiessen 22:23

Yeah, that's definitely an opportunity. Are there any other opportunities here for advertisers? I can't talk advertisers as a discipline. Like what yeah.

Mike Boland 22:33

I think just Generally speaking, whenever there's sort of like an upheaval, you know, industries get disrupted and everything sort of twists around and takes shape in like new shape. As that happened, gaps open up in the fabric, right? And I think that that's something that we saw happen in the past few big sort of tech revolutions, the smartphone and advertising as it adapted to the smartphone and media strategies and smartphone. I think that like those gaps are opportunities to be filled. Um and that might be an obvious statement, but I think that like sometimes people forget that in sort of avoiding the mistakes of the past few revolutions. Like a lot of the sort of the media, the media entities, I won't name any of them, but Celine, you and I've known these folks, you know, for the past 20 years in the sort of disruptions in the internet age and then subsequent sort of smartphone revolution. Um, those that sort of took a wait-and-see approach, or those that waited too long, uh sort of sort of missed the boat. And I think that, you know, that that happened in recent history, so everyone should remember that in terms of like acting fast. And I think that this the evolution of AI is actually moving a lot faster than the smartphone. And there are a few reasons for that, but one of them is it's it's it's bits rather than atoms, right? That these aren't, you know, the the the adoption rate of of the of the iPhone, for example, like it took a few years to get up to like 100 million users because these are talking about shipping actual devices and things like that. AI is software and it's moving so quickly and it's being adopted so quickly. So um I think that the opportunity to jump on top of it, whether you are a brand that's advertising, whether you're an entity in the ad stack somewhere, just it's gonna move quickly. So I think learning the the lessons of the past few tech revolutions is important in terms of not waiting on the sidelines. Um and I think that um that's all very general. I think to bring that to more of a specific level, let's take like the folks in in our world, which are media sellers. They could be um, you know, working in actual ad sales or in executive positions, whatever the case is, I think it's it's we we had Steve Pruitt on this podcast uh recently, um, who is from Cox Media Group. We were talking about automotive and we were talking about like good sales strategies. And something he said, I think really stuck with me, and I think it applies here, which is that as a sort of say it's from a rep's position, right? As a rep, if you just know your stuff really well and you know, go in to clients or potential clients and sell by not selling, but by being an education, a source of education, make them smarter about something that may be giving them anxiety, something new. And I think that that lesson applies here because there's a lot of anxiety around everyone that's just like this AI thing is happening so quickly, and I need to get up to speed on it so that I can like report to my boss on how this applies to us, or that I can, you know, just at least sound smart at cocktail parties, whatever the case may be. I think that as reps or as just media companies to educate their clients and be that sort of entity that holds their hand into this, you know, crazy new world, they will remember it. You know, you'll be their guy, you'll be their AI guy. And I think that that trust and that relationship will translate into dollars uh in terms of their spend uh on a longer term basis. So that's that's one sort of at least tactical piece is just like know it inside and out, spend your time listening. There's so many great podcasts. I started listening to one called How I AI. And a lot of great YouTube influencers that are doing educational stuff on like how to use Claude and just all kinds of cool stuff out there. It's free education. So like make yourself smart on this stuff and then be a consultative educational resource for your clients. And I think that that's sort of the best thing you can do these days.

Celine Matthiessen 26:35

Well, I think too, uh yeah, I think that's definitely you know the way to go because even even the you know, the media sellers that are out there, their companies are using AI or who they partner with are using AI to generate ads or to help come up with the audiences that when they plan campaigns, so they can, you know, honestly say, Well, we use it. You know, I'm already Yeah, learn by doing.

Mike Boland 26:59

Yeah, yeah, forget about the podcast that I mentioned. Learn by doing.

What OpenAI Does Next

Celine Matthiessen 27:03

It's funny that you mentioned the podcast because I have an email newsletter that I love um that gives me the daily down low and explains what's happening with all the different yeah, I'll send it off after this and put it to you, but it's really cool. And it also has like um it has fun prompts, like goofy prompts, like the goofy prompt of the day. Nice or like so it's neat, but it it's it can be fun. I think Mike, I think you're right. I think part of that is getting used to something new like this is you know, for the local sellers too, to just you know, be have fun with it, talk about your experiences with it, because it's not, I mean, we're all using it in some way, shape, or form. Um, you know, whether it be generative or, you know, you know, I'm not gonna say the A word here because it's in the office with me, but you know what I'm talking about. Um, voice recognition or voice command, you know, everybody has some sort of touch with it or have heard about it. So so final question, looking to the future, what once what do you think um open AI's next move is as the platform grows? Is it video product? What is it? What do you think, Mike?

Mike Boland 28:07

Totally. So they're sort of all over the place, but they need to be because things are moving so quickly. So, like I said, um, Claude is eating their lunch, uh, Anthropics eating their lunch when it comes to the enterprise. So I think that they're going to really put a lot of resources towards competing in the enterprise to catch up to Claude. One thing that they've already done pursuant to that gets to video, which you just said. So Sora, they killed Sora, um, which was their generative AI video generation engine. Uh, they're sort of leaving that market to a lot of the sort of standalone players that are doing interesting things with, you know, and and it was really computationally intensive. As we know, AI is operationally expensive. Um, so the video stuff was just really taking up resources, taking up costs, and they decided to put a lot of the those internal resources towards chasing enterprise. So I think enterprise is one. Two is, as I said, and all everyone's sort of speculating about potential IPO. I think, like a lot of companies there these days, they're they're putting that off with just like these just mega late stage rounds that are just just the the tech world has become known for, which allows companies, of course, to sort of put off going public and be under the scrutiny of quarterly uh reporting cycles and all the other headaches that come with being public. Uh, but at a certain point, I think they're gonna want to go public. So that probably is on their horizon. Um I think that um yeah, enterprise getting ready for um a public offering. Um, and and you know, they they continue to move so quickly. So I think that um, like another area is codex, which is there, it's basically the competitor to clawed code. It's basically the vibe coding tool. I think they're gonna continue to develop that. And I think also they're gonna start going vertical, right? So I think they'll start to roll out things that are like they'll choose these big meaty verticals and build purpose-built sort of um LLMs for them. I think they've actually already done that for healthcare. Legal makes sense for that. Yeah, that makes sense. Totally. Like so that um, you know, whether it be hospital administrators or doctors or nurses can have this really rich sort of um uh resource tool for them. And I think that some of those verticals have also regulatory requirements, like you know, HIPAA requirements in healthcare. So you sort of need an LM that's like contained within that world. So that's the reason they've gone there. I think legal is gonna be a good one for for sort of vertical drill downs in their innovation. So I think um segmenting vertically is another area they could develop.

Celine Matthiessen 30:43

Um and I think that you know you know what's interesting with both those verticals is they're both heavily driven by um uh by code. They both have their own codes. So if you think about healthcare, sure, all the all the diagnoses are codes, legal, same thing, which for an AI is very easy and very quick for it to learn because it's very black and white. There's not a lot of, you know, for if you had to like, you know, if you have to track some somebody and they're as a patient or as an ID in their um what they're going through and what they're prescribing and who they talk to, all that is, yeah. Um, and that's so that's really interesting. I just kind of thought about I thought about I'm like, yeah, that would be smart to start with, you know, with industries or verticals that are big, but also are very easy to understand how they're structured. Yeah. Yeah, and highly specialized.

Mike Boland 31:38

There's and there's you're right that there's a lot of there's a lot of structured data there. Um and and and LLMs like structured data. But also, maybe you know, this could be an opportunity to finally solve the decades-long problem of electronic medical records and sort of federating them and not having them be this like siloed mess. Maybe like there's some sort of agentic solution to that.

Next Topic: AI Search And SEO

Celine Matthiessen 32:00

Yeah, that keeps the privacy uh, you know, or at least gives me permission to provide the, you know, one one doctor to another without having to write, you know, sign paperwork to do that. And yeah. Yeah, for sure. Well, you know, Mike, this has been so amazing. Thank you for sharing your time and your intelligence and your expertise with us today. We I, you know, I always and we always appreciate and value your time with us. I do too, definitely. So, with that, I guess, you know, we have another one coming out. Should we talk about what we might be doing for the next one for those listening? What did we say we were gonna do for our next one? Was it on AI? We were gonna do another one on.

Mike Boland 32:45

Yeah, no, I think this this, as as we just talked about, it's moving so quickly. So I think this will be sort of an ongoing topic that we'll cover. Um but but one sort of one sort of sub segment of this overall AI discussion that I'm excited to talk about, which uh we can sort of tease for our audience, is the the intersection of AI and search, which we talked about in this episode, but particularly in the art of SEO. You know, the world has been sort of learning and mastering the art of SEO over the last 20 years, and now the whole thing is turned on its head because LLMs have sort of different ranking factors in terms of the things you do to make sure that your site will surface high in Google results. Um, now some of that is inherited by LLMs and how they sort of surface businesses and make their recommendations. Say you're looking for, you know, a dog washer and you go to Chat GPT to ask them. You know, it's going to be slightly different in terms of the playbook for gaining visibility in these engines. And some of it, again, is the same, some of it's different, and and the SEO world is trying to figure that out. We've been looking at it very closely. So I'd love to break that down for our audience in in an upcoming episode.

Celine Matthiessen 33:57

All right, audience, you heard that coming soon. So, and any questions you have for Mike before we record it is um you can email us podcasts@ BIA.com. So, for everyone out there listening, thank you for joining us as well. Um, we're hope we hope you're enjoying our podcast this year. And like I said, if you have any questions for our next one or any topics you want us to cover, especially, you know, since we've got Mike uh on call for podcasting for for now, right? Yeah, send us an email at podcasts@bia.com. And we look forward to bringing you more insightful conversations on local media throughout this year. Thank you.