Loading summary
Transcript88 lines
- [00:00]
A
Foreign.
- [00:09]
B
Hello, hello and welcome to another episode of the Digiday Podcast, a show for people who have asked ChatGPT, did a brand pay you to say that? I'm Kamiko McCoy, senior marketing reporter here at Digiday.
- [00:22]
A
I'm Tim Peterson, executive editor of Video and Audio Digital Media. What's up, Kimiko? Happy Thanksgiving week.
- [00:28]
B
Happy Thanksgiving week indeed. Do you have any interesting plans for Thanksgiving?
- [00:34]
A
Yeah, so I'll go up to LA to spend Thanksgiving with my family. And. The people who host Thanksgiving have a history of saying, oh, we're going to eat at 3 o' clock and it's not until 6 o' clock that you eat. So I think last year was the first year we started this tradition where my significant other and I go to IHOP with my dad to make sure where you've eaten before the actual feast. What about you?
- [01:05]
B
You know, I'm in the same boat. Ours usually starts at four. I think everybody's lying because we don't eat till six. I feel like four is when they start, like the cooked turkey's done, but that's when they start the sides and I've been bamboozled. So, yeah, absolutely. I. I like to think of like a nice French toast breakfast to fuel me for the day before I am inevitably eating it. So 6:30. I also believe in to go boxes. Bring them yourself and take Thanksgiving leftovers with you home. Okay.
- [01:33]
A
Yeah, boxes is smart. We usually do Ziploc bags, which is a choice, but it gets the job done. Pie is a little questionable, a Ziploc bag, but I'm a bowl person. Just like throw everything into a bowl and give me a spoon and I'm good. So Ziploc bag is just kind of like a to go bowl.
- [01:53]
B
Yeah, a Thanksgiving slop bowl, if you will.
- [01:56]
A
Oh, God. Oh, and speaking of AI slop. So for this episode, for our feature segment, we had on digi platforms reporter Crystal Scanlon to talk about when are OpenAI's ads coming? What's going on there? Especially because it seems like as we're talking about with Thanksgiving day after Thanksgiving, Black Friday into Cyber Monday, kind of potentially unofficial kickoff for that ads business, based on some recent comments that OpenAI I execs have made. But we get into that with Crystal later in the episode. First we got to talk about the news because there's been a good amount of news coming out ahead of the this holiday weekend.
- [02:38]
B
Yeah, we've got our jazz fingers here. Juicy scoops. An update with Warner Brothers discovery bids that are coming from anywhere and everywhere. Meta's antitrust case and how that went, the EU Commission on the Omnicom IPG deal, and also Adobe and SA SEMrush.
- [02:56]
A
Yep. So let's start with Warner Brothers Discovery, because that, See, there's just so much activity there. It's the new Paramount in a way. So Skydance finally acquires Paramount, and almost immediately new owner David Ellison says, cool, what can I buy next? And Warner Brothers Discovery is the one on the block, which seems fitting because Warner Brothers, WarnerMedia at Time Warner, all its previous permutations, is always getting acquired or getting sold or getting merged or getting decoupled. In this case, Warner Brothers Discovery has already said, we're going to split into two companies in 2026. One's going to be the studio and the streaming business. The other is going to be the cable TV network business. That kind of no one wants to be in anymore. And this is where new Paramount owner David Ellison said, cool, can I have both, please? And now Comcast and Netflix have come in. So last Thursday was the deadline for bids to be submitted. And as you mentioned, Paramount, Comcast and Netflix have all submitted bids so far. So these are the three potential suitors for Warner Brothers Discovery, but they want different things.
- [04:19]
B
Yeah, that's. And also, not for nothing, Paramount has asked three times and have been turned down by Warner Brothers Discovery three times on a matter of, you know, is the price right? So it'll be interesting to see, you know, who gets it in the end. But yeah, besides Paramount, which is asking for everything, comp.
- [04:39]
A
Give me all of it, please.
- [04:40]
B
Take the Comcast. I want the entire sloppole.
- [04:44]
A
Comcast and Netflix, Yeah, that's the media business.
- [04:48]
B
Just throw it all in there. Comcast and Netflix want just the studio and streaming business, whereas to Paramount wants all of it. So it'll be interesting to see kind of if and how that deal shakes out.
- [05:00]
A
Yeah, Comcast, Netflix is saying like, no, thank you on the Discovery Networks.
- [05:06]
B
Yeah, keep that.
- [05:07]
A
No, thank you on the Turner Networks. No, thank you on cnn. We are good without that news business. We would just like the film and TV studio because we're already in those businesses. Those are good businesses. And we would like the streaming service because now hbo, Max, that's a good brand. Which is interesting because, like, HBO was historically Netflix's primary rival for a time, where I think it was Reed Hastings, former CEO of Netflix, who said, we're trying to become HBO before HBO becomes Netflix. They're past that now. They're worried about YouTube. That's their primary rival. But it seems like they would like to have HBO as to help with that competition.
- [05:52]
B
That does make a lot of sense. And like I mentioned, it'll be interesting to see how it shakes out because like if the deal doesn't go to Paramount right then what then happens to the redheaded stepchild, which is their cable TV business, because it's still going to be split into two.
- [06:04]
A
Yeah. Yeah. So it seems like maybe that ends up getting paired with. I still don't know how to pronounce it. Versant. Versant. I've heard people say it out loud so I've been exposed to the correct pronunciation, but it's never stuck. But that could be one option. It seems like there's going to probably be some roll up of cable TV networks as we've seen with like local newspapers and like kind of even some digital media properties as well. But so these three birds are in unclear when exactly a decision is going to be made. I think David Zaslav, the CEO of Warner Brothers Discovery, has said they're looking to be making some decisions right before the Christmas holiday. So I think it was like around December 20th timeframe, but I don't know what exactly that decision is supposed to be. So it probably won't be until 2026 when we get any sort of verdict here. We did however get a verdict this past week in the Meta antitrust trial. So the US Federal Trade Commission had sued Meta claiming that its acquisitions of Instagram and WhatsApp violated antitrust law. Judge last week said, nah, not proven unstable.
- [07:28]
B
A big get for Meta here. And I almost feel like it was falling in the same veins because like when that was first announced, it was around the time that like Google was making having a decision made on its business and whatnot. So you had all of these tech titans that were being questioned by the government and all of them have made out unscathed.
- [07:47]
A
Yeah. Although we'll see on the Google Ad tech antitrust case because, oh yes, still pending week was closing arguments, but I think it's not going to be until early next year that there's a verdict on that one. So. But based on like this Meta case, it seems like the tech companies are fine. Meta's case got helped by TikTok, the rise of TikTok and the continuation of TikTok's availability. Now it's one of those sliding door moments where if TikTok had got banned in the US year, how would that have changed things? But because it's continued to be available and has grown, that seemed to be one of the major reasons why Meta is able to now hang on to Instagram and WhatsApp.
- [08:33]
B
That's really interesting. And like, even when the question was first bought up, brought up, excuse me, about it being an antitrust, because Meta's thing has always been, Zuckerberg's thing has always been like, if you can't beat them, buy them, which kind of happened in this case. And to see a judge say, well, that's unfounded, unproven, you're good to go, you know, collect the money, pass, go. Says a lot about kind of where tech companies stand and how, how much they're able to control.
- [08:59]
A
And does this now open the door to tech companies to get acquisitive again? Because it seemed like because of the Google antitrust case and this Meta antitrust case and also just Lina Khan, when she was running the FTC and was very much critical of the big tech companies, whether all of those concerns about, like, if we're a big tech company and we try to acquire another big tech company or a big ish tech company, are we going to be able to get that deal through? Maybe, because deals are getting through big deals. So as you mentioned at the top, the European Commission has approved the Omnicom IPG merger without any conditions. And so this now clears this merger for acquisition. I think it's more technically like Omnicom's acquiring IPG because of the size differences between the two. But so this clears the deal in European Union. It's already gotten clearance in the United Kingdom and it's already gotten clearance in the United States. So seems like all signs go for this deal to close maybe by the end of this year.
- [10:08]
B
Yeah, I think it's like 48 hours that you expect to hear some more news afterward, if I'm not mistaken, as sort of that turnaround time. But it'll, it'll be interesting. And I know that we've got this coming up on a podcast further down the road with our senior marketing reporter Sam Bradley about kind of like IPG holding companies, WPP and kind of their waning. Right. And kind of how agency hold codes are kind of to find a way to make themselves indispensable, I think by clearing this deal and having, I don't know, more tools in your tool belt, a bigger presence and things like that could definitely be helpful. But I. It'll be an interesting kind of way to see how that shakes out mostly in WPP's case as they've got mergers and acquisitions happening on the other side of things. But also, speaking of mergers and acquisitions, did I beat you to the punch with Adobe and SEMrush?
- [11:00]
A
No, I was actually going to like, say what's going to be interesting, you know, when it comes to oupp. And we didn't really get into this with, in the conversation with Sam for next week's episode, but anytime you have a big merger or acquisition like this with Omnicom IPG, it's. It takes 12 to 18 months for that company to go through the integration phase. And so it's, I feel like historically people have said, like, companies are kind of put on ice during that integration. Like, there's only so much that they can be doing because so much of the work is just like getting teams together, figuring out like back office stuff. And so I wonder if that could create a bit of a window for WPP to rebound. But it's got a lot to rebound from as we talk talked about with Sam. But we'll have to see that for next year once this deal does close. Another deal closing. Adobe acquiring semrush. So semrush is like a search analytics provider. It's one of the companies that a lot of publishers as well as brands are using to figure out, hey, how's my brand appearing in ChatGPT or Gemini or the other AI chatbots? To what. What are people asking about me? Or like, to what extent are these different AI chatbots citing me in response to people's prompts? And so it's paying $1.9 billion for this, which is a pretty penny. And it seems like the idea is to integrate SEMrush into Adobe's existing marketing tools. So giving brands a better idea. The brand's already using Adobe software, a better idea of how they're appearing within AI chatbots, which seems like could prime the pump a bit for when OpenAI finally decides to get into the advertising business. Which is our topic of our conversation with Crystal Scanlon for this week's feature segment.
- [12:59]
B
Absolutely. Yeah, we touch on that and kind of put it all into context. I think Crystal's got great big brain energy to kind of walk us through all of this and what this means giving. She's been covering this space so closely, but I do think there are a lot of questions that marketers have around kind of how their brand appears, messaging and like how they can and I don't want to use the word hack, but hack these systems. Obviously having that integrated tool helps. But yeah, the conversation with Crystal, juicy, in my opinion, and kind of answers a lot of those questions based on what she's reporting.
- [13:29]
A
So let's get into that. And Kamika, enjoy your Thanksgiving.
- [13:33]
B
You too.
- [13:45]
A
Crystal. Welcome back to the show. Is this your third or fourth time on the show this year?
- [13:50]
C
It's my sixth. It's my sixth.
- [13:52]
A
Wow, what a year.
- [13:55]
C
I'm becoming like a regular.
- [13:57]
A
Well, thanks for joining us. So we're talking about OpenAI today and this looming ads business. OpenAI has kind of taken the baton from Netflix in terms of being the the next big ads business everyone expects to drop at some point. Kimiko Krystal, I want to get your take on a kind of theory, I guess you could call it that. I have. I feel like this Black Friday and Cyber Monday is the unofficial kickoff to OpenAI's ads business. Now I'll explain why I feel like that's the case. So there's this Wired interview that just published in which Fiji Simo, who's the new CEO of Applications at OpenAI, basically is the person in charge of the consumer product and consumer facing revenue at OpenAI for ChatGPT. They ask her about, hey, when's an ads business coming? How are you thinking about advertising? And she said, quote, advertising as a model works really well when you have a lot of commerce intent. And it seems like OpenAI is doing so much right now to establish that commerce intent. And obviously people do a ton of shopping around Black Friday and Cyber Monday, but Modern Retail, our sibling publication, Mitchell pardon Over there, just reported on how Target is rolling out a ChatGPT app ahead of Black Friday for people to then be able to do their Black Friday Cyber Monday shopping through ChatGPT to target Walmart, seeing a bunch of referral traffic from ChatGPT. It's also part of this instant checkout program, as is Shopify, as is Etsy. So it seems like they're putting everything in place to establish shopping through ChatGPT. So then they can go to companies and say, hey, you're getting a lot of sales through ChatGPT. Would you like more than you should pay us for these ads that we have not defined yet? Is this an outlandish theory or is there something to this? What do you both think, Crystal, maybe start with you?
- [16:06]
C
I'd say, yeah, I buy. I buy that personally. I mean, even I read that same interview and literally for me, it definitely came across like they need to have three things in order and before they can look to have an ads business, it's the issues around data privacy because they do have a wealth of data under their belts. And obviously that's going to be what advertisers want. They need to have the commerce proposition fully netted out and like really working well. And then obviously once they've got those things in place, which are kind of the material ideas, they need the guy or girl, woman, whoever to be able to sell those ads. So I'd say yeah, this Black Friday, if all goes well, I'd say it's probably going to fast track it, potentially.
- [16:47]
A
Kubeco, you're generally the realist when it comes to AI on this show. What do you think?
- [16:53]
B
I think it's a fair hypothesis and a fair theory and I think it'll be a good indicator. You know, people are expected to spend a ton this year even with economic headwinds for like Black Friday and Cyber Monday. Right. And then I will always go back to the idea and they even asked Fiji during the interview. You guys, they're burning through billions. They've got to find a way to kind of make up those costs somewhere. Why not start during a guaranteed shopping period?
- [17:21]
A
Kamiko, are you planning to do your shopping through ChatGPT?
- [17:25]
B
Absolutely not.
- [17:29]
A
Krystal, what about you?
- [17:31]
C
I mean, I'm going to be in the US During Black Friday, so I'm going to be doing it in person rather than through an app, albeit I will probably still be using apps. But to ship to London, bring back.
- [17:45]
B
The days of us all fighting for TVs in person. Why ask an app when you can fight over a $20 TV at Walmart.
- [17:54]
C
America's pastime, we'll say Black Friday is a special occasion from what I've seen there, that's for sure.
- [18:02]
A
Although I wonder if that's. Well, we'll have. So for our sibling podcast, the Glossy Glossy Beauty podcast and the Modern Retail podcast, they'll have a special episode on Black Friday and Cyber Monday Shopping. I'm very curious on like to what extent people do much shopping in store on Black Friday anymore. I'm not a big shopper, so I'm not case study of one. I'm not going to be representative. I am very curious though in to what extent people do shopping through something like ChatGPT. Again, kind of coming back to this hypothesis is the word that I couldn't think of. Kimiko, thank you for putting that back in my lexicon. This hypothesis of Black Friday Cyber Monday kind of being the unofficial kickoff to the ads business. But Krystal, you made the point of like, okay, yes, step one, it seems, according to Fiji Simo, is they gotta establish that commerce intent. But there's so many more steps after that to actually building this ads business.
- [19:05]
C
I mean, yeah, like I guess for one, I suppose the kind of the only thing that I've really seen or really the main thing that I've really seen if we're gonna as in like I guess a prelude to what they could or couldn't do is what we saw with perplexity and what kind of went wrong with that. I say wrong. Wrong is potentially the wrong word, but it's wrong in the sense of it didn't really seem to spark much interest from advertisers bar the initial launch. Outside of that initial launch of a handful of advertisers, there was never any more announcements from advertisers like of advertisers taking part in it. Any that I spoke to were having trouble to get onto the whole scheme that they were doing. And a thing that was a real sticking point for advertisers as well was the fact that they decided to charge for the ads around CPMs rather than CPC. And that was just a whole like mind game situation that advertisers couldn't get on board with because it's not the way that search is normally bought. And they were like, well, if you're coming into a search world, why would you not do it in that manner? Perplexity's answer was well, we're not trying to do what was already done, we're trying to do something different. The problem is it's not a tried and tested platform so it didn't really work out well for them. And obviously since then we've seen their ads. Chief Tas Patel, he's stepped away completely. He's no longer in the business. It seems like they've potentially, it's on a bit of a hiatus for now. So I'm guessing at least if we're looking at that as one example of what not to do. I'm sure Fiji's definitely seen it and just that's definitely a lesson of let's not go down that path.
- [20:37]
A
No, but it's interesting because this is also going on at a time where we've seen this huge boom in retail media spending, commerce media spending writ large because it's not just retailers but it's airlines, hotels, gyms. Kimiko, you've been leading our reporting on this and so I wonder if this is something where this will work in OpenAI's favor that there is so much new commerce media inventory and the idea of historically you were using search advertising if you were trying to drive online product Sales. Now you can be using not just Amazon, Walmart, Target, but using that data in connected tv, digital out of home. Whether there is something more of an openness from advertisers to be buying different types of inventories. So long as the commerce intent and the commerce data underpins that, that which I think is the thing that Perplexity wasn't able to establish first.
- [21:43]
B
I think that's fair. And then also if there was anybody who was capable of doing it. You have to look at Fiji's success with Instacart, hence why she was brought into OpenAI. Right. So the roadmap has already been created and if you look at Instacart's retail media, they've been one of the top players in this space. So I think if she is able to take the tools that she built and the team built there, I think it's easily transferable and they may be able to stand as kind of the canary in the coal mine in comparison to something like Perplexity.
- [22:14]
A
What's been the legacy of Instacart's retail media business? Because like, I'm aware that they're one of the bigger players in the retail media or commerce media space. That said, I still feel like I hear less about them than like Amazon, Walmart, even Target, Uber. And I don't know if that's just like a PR thing. I don't know if that's just because I don't use Instagram. I don't know if that's just because we're talking about Amazon, Walmart, Target, like these huge companies.
- [22:46]
B
It's the latter. And they also don't have like a brick. The thing is like you've got. I'm getting a little off beaten path here, but the retail media networks that have a physical retail location have a little bit more of an edge around like some of the data and some of the closed loop and things like that. Now Instacart has been really competing in that place with their own tech stacks and partnerships and deals that they've been inking to kind of like get more data and feed the beast there. But I mean, Amazon and Walmart and Target compared to Instagram, those are just like really big behemoths. It's like a, you know, a Google versus Bing situation. Not to say, you know, Bing is not killing it in their own way, but comparing them to Google is a. That's a giant.
- [23:28]
A
Okay. Yeah. Because one of the reasons why I ask is to what extent that gives us kind of a. What is it Rosetta Stone into okay, what could CMO be doing at OpenAI? Because it seems like a challenge OpenAI is going to have is what should the ad product? And Crystal, when we had you on the show, what was it like a month ago, I guess your fifth time this year? On the show we talked about what should the ChatGPT add product? B Search Engine Land reported on an interview that OpenAI CEO Sam Altman had with conversations with Tyler in which Sam Altman seemed to be, like, very candid about having no idea what the ad product within ChatGPT should be, other than he seems to think the Google, the core Google search ad product is actually a bad product in a way, because he says something to the effect of if you're having to show an ad for people to click on to get to what they want, that means you weren't able to organically give them the information that they would want. And then he kind of gets into this idea of, but so if we could have chatgpt kind of like organically give results of, if you're going to New Orleans, where should you go for a restaurant? If you click one of the suggestions and book a table with that restaurant, we should get a cut from that, which becomes a pay to play thing. We don't have to get into the conversation about, like, isn't advertising just pay to play? But it seems my takeaway from that conversation or search engine recap of that conversation is it's going to be really challenging for them to figure out what this ad product is meant to be.
- [25:20]
C
Yeah, I mean, even anything I've read about what Sam Altman said, like, the only ones where he's kind of been somewhat, I guess, complimentary about it was literally Instagrams. And that's because he's admitted himself, he's bought from them. So he was like, they clearly work because he's literally bought from them. But Instagram itself is also a very different platform to ChatGPT. So it's definitely, it definitely seems like it's still very much in the early days of, I mean, what on earth could they do? Like, how could they, how could they not make the same mistakes that everyone else has already made? But also what he's definitely made a big thing about is that he doesn't want to compromise any trust either with the users, and he doesn't want them to feel like what he sees on other platforms at the moment is that the ads aren't necessarily well integrated. And that's half the problem. So it's how, how do you even do that in a place like ChatGPT where, I mean it's still text based predominantly. Unless you ask it to give you an image of something. It's not an Instagram type platform. But then how do you do that? Make sure that it still integrates well within the user experience without literally, I guess, annoying half your users. Because then potentially if you annoy the users, you might compromise the subscription play. How on earth they do that. But I guess again, that's probably also why Fiji's in there. Like she's literally been at Meta as well. So she's learned from the best ultimately. So she's kind of got their playbook in her mind. She knows what worked well and what didn't. She's literally had probably like Zuckerberg on speed dial. So in that. Not that he's giving her his trades.
- [26:53]
A
I was gonna say, I wonder how much they're actually talking these days now that they're going real head to head.
- [26:58]
C
Well be like, hi, how are you? But actually, don't tell me you also.
- [27:02]
B
Have to consider like, I don't know, she's got a big task ahead of her because at Meta and instacarp, the, the path to success seemed a little bit more clear than it did with like an open AI. Because like even in the Wired interview, the first line is that they've got a weird structure because it's a non profit in charge of a for profit that's become a public benefit corporation. So like even as open AI kind of struggles to define itself right. It's hard to then carve out a path of like what your ad products can, can look like. And they've got a suite of like, I don't know, to me it kind of seems like throwing spaghetti noodles at the wall to see what sticks. Even with their like products like Pulse and some of the other ones that they've, they've put out there to see what people gravitate towards. Sora, all of those.
- [27:46]
C
Yeah, I suppose to be fair, when I did actually meet with them a couple of weeks ago, they did openly say, well, openly open AI, I get. They did definitely say to me that when it comes to product launches and stuff, like they can tell me categorically today they don't have anything. That doesn't mean to say that's going to change. Like, that's not going to change like literally overnight because it literally depends on what team's working on what and what's ready to ship at what point. There's not really like A categorical we're doing this by this date, this is coming out on this date. There's not really a pathway they are working towards building that type of thing. But I guess that probably makes sense then why we see things come out so sporadically.
- [28:23]
B
So how do you build an ad product? Like how do you build an ad business around that? That's chaos.
- [28:28]
C
I guess when you think that they do have a lot of like they're kind of consulting with a lot of brands and advertisers at the moment like from a consulting perspective like talking about their compute and trying to help them be able to work better behind the scenes internally with their own teams. So already if they're forging those relationships, they're learning what those type of businesses want if they're going to be selling to them ultimately because they're not going to go into it blind because they've seen how other businesses have crashed and.
- [28:53]
A
Burned with is interesting that like looking at who they're working with for instant checkout who they're working with ChatGPT apps. The retailers who are in there are Walmart or Target or Uber are I think booking.com and or Expedia also have commerce media networks. Basically these are retailers or you know, commerce companies that have ad businesses already. It feels like that could be a way for OpenAI to start out. It's ads business standing on the shoulders like basically just Target, Walmart, add our inventory to your mix and that's how we can start this thing. As opposed to the perplexity problem of trying to come to the market with something brand new.
- [29:43]
B
Yeah, bespoke.
- [29:44]
C
Yeah, it definitely makes sense because I guess then you've already got the trust of those big brands so they're not going to allow some random effectively to just come and position themselves on their sites or their app. So if you're already going on the trust there then it's almost like you're already given a leg up regardless.
- [30:03]
A
Yeah, and it seems like the incentives would be there for the commerce media companies too because Kimiko hasn't the big trend this year especially been commerce media inventory expanding beyond like the owned and operated. So it's not just Amazon is like an early and best example of this but like all the others being like okay it's not just going to be ads on our commerce sites and our search results but how do we get into connecting connected TV out of home, things like that.
- [30:33]
B
Yes, precisely that. They want to create an ecosystem where you can really connect like follow the consumer journey, connect the Dots and things like that. Whether it be like, I think Best Buy had made a play for its take in store takeover to connect that with its digital play and whatnot, which was like a really big. Mitchell covered that as well, which is a really big deal to kind of flicking at your point.
- [30:55]
A
So. So it's going to be. And then I guess they still have to figure out that ad product though. But there was an interesting new wrinkle in all of this, it seems. So Google just launched Gemini 3 and it feels like every time one of these chatbot companies or AI companies launch a new consumer product, it kind of forces everyone else to follow suit. We saw this especially with Claude code being ascendant. And so then OpenAI and Google really tried to step up their coding agents. But the big thing with Gemini 3 seems to be its multimodal responses. Dumb word that basically is just like. It's not just text that it's thrown back at you, it's not just images that it's throwing back at you, but it's basically throwing web apps back at you now where you could be putting a prompt into Gemini and it'll create a web app as an answer for that. And so then it has visual capabilities in there and it feels like there would be a way of if ChatGPT wants to go in that direction, then you can have interactive ads in there and maybe figure out how to like bring advertisers in in a way that's a bit more elegant than just here's what you should buy or here's a sponsored product that you should buy. And we're just going to have like some words and a text link to it. Like it goes beyond the core Google search text in a link ad format.
- [32:35]
B
Good. I think marketers, the ones that I've talked to, are still scratching their heads trying to figure out how to even show up in these environments. They're also throwing spaghetti noodles at the wall, trying to like do a lot of traditional marketing tactics to show up in these environments to then determine how dollars could shift and kind of get. Get it to be a sure thing as opposed to a testing and learning.
- [32:59]
A
It does seem like they're starting to get a better idea of how they're showing up. So when we had Alaina McGurn from Digitas on the show, when Kimiko, you and I interviewed her at the Digiti Programmatic Marketing Summit back in May, she talked a little bit about how they're able to get some data from these various chatbots. I did an interview with Forbes during the publishing summit in Europe in October that Digiday held. And Nina Gould, the Chief Innovation Officer there, talked about how they're getting actual prompt data with referral traffic that they're getting from platforms like ChatGPT. And I think Sam Bradley's been doing a good amount of reporting for Digiday on this. Sarah Guaglioni has been doing reporting for Digiti around this. But it seems like with companies like SEMrush and similar web, that data is coming to companies in terms of how am I showing up in ChatGPT results. Also, these companies are seeing traffic from ChatGPT. Like I think Mitchell at Modern Retail reported this, that one, this was in August of this year, that one in five of Walmart's referral clicks came from ChatGPT, which is just a. A huge share of traffic to Walmart.
- [34:20]
B
Yeah, it is. And then kind of upends the. Well, I mean, we've talked about this already. Kind of upends the search, like how a search happens, what becomes the first one. And then that opens the door and brings us full circle to commerce, media and where not Instacart, where OpenAI and some of these other AI platforms could fit in.
- [34:38]
A
Yeah, because it feels like companies are getting at this point more than a glimpse of how ChatGPT benefits their site traffic, for one thing. And now, so we're recording this On Wednesday, November 19, Adobe has announced it's looking to acquire Semrush, which is really interesting. I wouldn't have expected Adobe to acquire semrush. For anyone listening who's not familiar, semrush is basically like an analytics provider. And one of the things that they're providing companies is data like we've been talking about in terms of how brands or publishers are showing up in ChatGPT conversations. Adobe buying SEMrush is interesting because there is like. It's interesting. It's weird. I'm still trying to make sense of it. I just saw this news a couple hours ago and I haven't been able to spend too much time thinking about it. But like a lot of brands use Adobe products to create their ads. So there is kind of a through line there that said, I don't know how. Adobe also has the marketing cloud. There is an analytics suite in there, but it's not the same thing as like a measurement provider buying SEMrush.
- [36:01]
B
Well, I mean, this is a hypothetical, but to me it points to. Adobe is maybe seeing the writing on the wall here. It's the same thing with Like, I think I touched on this before mergers and acquisitions in the influencer marketing space, where you had these big hold codes go to go buy up like tech plat, like influencer tech platforms, seeing where the dollars are going. And I imagine that if Adobe can kind of see the writing on the wall of, like, how people are shopping, how that's changing, and if they're looking to like Power Ranger style, build out their own offerings, as opposed to building something bespoke, you know, scoop up SEMrush and now you've got another tool in your toolkit.
- [36:38]
A
Yeah, because I'm trying to, like, I guess the thought exercise I've been trying to go through with this is, Does Adobe buying Semrush, is that another catalyst for OpenAI's ad business? Not that it would be like exclusively benefit OpenAI, but if companies who are using Adobe products to create their ads now have access to analytics in terms of how people are using ChatGPT to ask about those brands, does that give brands a better idea of how they're showing up in ChatGPT results in the context in which they're creating ads? Does that spur things along in any ways? Because one of the things you always hear when it comes to any new ad product or channel type is agencies talking about like, well, we need to do a lot of client education. We're doing a lot of client education. The most important thing right now is client education. Adobe owning SEMrush seems like it could help with client education, if only by making this data in some ways more available than having to do a deal directly with SEMrush. If you already have an enterprise deal within Adobe, I don't know, maybe I'm just spinning my wheels out here and what I need is a break from all these things. I am curious for listeners to comment on the show what they think about this, but Krystal, have you thought about any significance of this deal? I know it just broke this morning.
- [38:16]
C
Yeah, I'm still trying to wrap my head around it and just who's buying the who situation right now. But I mean, I guess this weird analogy, not even an analogy really, but considering you've got other companies already partnered with OpenAI, partly because as well, they're wanting to be. They want to be seen as being at the forefront of AI because they can see that that's going to be. That's the future. We need to get on board with it. We need to figure out the best ways to use it. In some respects, OpenAI is giving them an engineer of sorts to be able to get up to scratch with the different tools, figure out how they can work things internally. So in that sense, I guess I kind of see it in a weird parallel way. So like if you've got Adobe doing all like the creative side, but then you've got this data, it kind of gives their, how do I put it?
- [39:06]
B
It's.
- [39:08]
C
You've got OpenAI helping the companies learn how to use all the AI tools so they're all up to scratch, but then at the same time then you've got the data to know how, what they need to do to be able to adjust themselves so that then they show up right in ChatGPT. That's about as far as my head has been able to literally try to digest any of this right now.
- [39:27]
A
Yeah, no, because I feel like what the three of us are doing right here is like tea leaf reading, but looking at different cups and just like, okay, how do we understand the teapot that this all came from based on the leaves in these various cups? I sure think this is now definitely by far the worst analogy on this episode potentially. So I'll stop with it. But it is because, I mean, another thing to be thinking about is there was the Wall Street Journal article from, I believe earlier in November in which Veren Shetty, who's the head of media partnerships at OpenAI, talks about wanting to have brands like McDonald's make their mascots like Ronald McDonald available for people to create videos within Sora using. Which I think is something we had talked about Crystal, when you were on in October and Sora had just launched. But the idea of sponsored cameos or branded cameos, it seems like OpenAI is like, yes, absolutely, we want to do that.
- [40:31]
C
I wouldn't even, to be honest. I mean, I don't know if this is far fetched or not. We all can see that there is definitely a struggle with how to make ads really integrated in ChatGPT. I mean, considering how they're very sporadic in terms of what they release, when and how they release it, who knows if they then just not necessarily can chatgpt ads. They might just start off with Sora because it's a format that we already know that Sam Altman likes because he likes Instagram ads. A lot of the Sora content is already showing up in Instagram. So in that sense it's kind of a format that they're all very familiar with and it's kind of almost trusted ground in that format. But then I'm also potentially going off on a Very wild tangent when we know that ChatGPT is what they ideally want to do first.
- [41:17]
A
And as Sora, what's both of your senses in terms of like, has SORA been able to keep up the momentum? I feel like it's died down precipitously.
- [41:29]
C
It's definitely from our side, it's definitely died down, but I think also it's still only invite only as far as I know. From what I understand, even coming to the uk, I was definitely told that it we shouldn't be having to wait long, albeit they wouldn't give me a timeline. But I mean in that respect, I think considering they're still in vite only, it's done incredibly well. But I'd imagine it will probably have another peak again when they open the doors or open the floodgates because then it's like there'll be a big announcement about it. Everyone that couldn't get in probably has FOMO and then wants to see what it was all about. And that's when we're probably going to start seeing a lot more of the data issues. Because then what the issues we saw initially around the MLK content and other kind of questionable content, there's bound to be an influx of that at the same time. So it's how do they then balance sorting that out with a lot of more of the tools and restrictions and guardrails around that with trying to build an ad product around all of these different things and then try and keep advertisers happy so that they're actually okay with the thoughts of their ads being around any of this type of content on any of their platforms.
- [42:36]
A
Kamika, have you gotten on Sora or you haven't any Sora fomo?
- [42:40]
B
No, absolutely not. No. I'm in the AI slop camp, don't like it. But I think, I mean it takes us back full circle where I think the reason sometimes like part of it with OpenAI and Fiji and Sam and all of the other execs there, why they're so vague sometimes about ads and what those ads could look like or their ad plan is because like I said, there's so many moving parts here. You could put ads in Sora, you could put monetize, ChatGPT, but a lot of it, there's a lot of unanswered questions around safety hallucinations. Even when you think about like Sora and brand deals, there's that spark conversations about like copyright and IP and things like that. So I think there's a lot of things to answer before they can start properly rolling out what ads look like for them.
- [43:34]
C
I did quite like that interview though, because the fact that she was just transparent about that and just not just trying to make up the fact that they know exactly what they're doing with it, that they are trying to almost take a measured approach. Even if a measured approach ultimately means like they do a bit of a something's ready by next week and then they just run it out to market. At least that seems from this initial stage there is thought behind it and how they're going to do it. Because I mean Fiji herself has worked at like Meta. She knows how these things can work and how they can work really, really well. She's been the CEO of Instacart like she's tried and tested in herself, having had success behind her. So I guess at least we can be hopeful that she's been transparent on that front.
- [44:13]
B
Can be.
- [44:15]
A
So we started with Black Friday Cyber Monday being kind of one potential tell for the open eye ads business. I feel like the next one after this will be if Sam Altman or Fiji Sima or probably most likely Varun Shetty are at CES taking meetings with brands and agencies.
- [44:39]
C
So when I did meet them they said they weren't doing ces. Whether that changes again since things seem to change there a lot is tbd. But they did say at the time they weren't going to be there. One place they did say they are considering and already in talks about is Cannes next year. So to my mind is you're going to want to have some form of product before you show up there. Because if we kind of compare them again to perplexity. Perplexity was at Cannes this year, but it was very like under the radar, very quiet meetings. I think there was one lunch and it was still the message was coming from those individuals as ADS is not really going to be like a real money maker for at least the next five years. Like we're still putting our feelers out. What does the market want to do? It doesn't seem like OpenAI would ever take that sort of approach. They're doing all of their groundwork and foundational work now. I mean I've heard in some conversations that people would be surprised if the name of an ADS boss wouldn't be mentioned by the end of this year. Or we at least if we don't have a confirmed person then at least some kind of front running names as to who the kind of potential options by the end of this year they would want to definitely have that in place because that person is going to need some time at the company before being able to go straight out to Cannes.
- [45:54]
A
Unless they want to do the Netflix Playbook, which is announced in March. We're doing ads. We have this external partner that we're working with to stand up the ads business. Big splashy meetings at Cannes and not until after Cannes, hiring the ads boss who then gets similarly let go a year or two later. Two years later, I think it was.
- [46:20]
C
I mean, yeah, I, I, at this point I'll buy any of these options because it's, it's still very much of a tbd. But I mean, who knows at this point if they go down that route, they might even be doing hope, holding out interviews again.
- [46:35]
A
So we'll have to see. I'm sure. You know, Crystal, we've had you on six times so far this year. I'm sure in 2026 it'll be more than six. Maybe it'll be 202026 for 2026. But thanks so much for coming on the show. Really enjoyed speaking with you about this.
- [46:55]
C
Thank you both for having me again. Always a pleasure.
- [47:04]
B
Well, that brings us to the end of this episode of the Digiday 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 Podcasts. We'll be back next week with another episode of the Digiday Podcast. Thank you so much for joining us.