← Mobile Dev Memo Podcast
Mobile Dev Memo Podcast
Season 7, Episode 13: Google navigates the AI advertising era (with Dan Taylor)
00:46:06
Loading summary
Transcript98 lines
- [00:02]
Sponsor/Ad Reader
Branch is an AI powered MMP that helps growth teams see exactly where to put their dollars to bring users in the door and keep them coming back. If you haven't had a chance to check them out, now is the perfect time to do it. They've recently updated their look and feel to reflect the innovation that's happening in their product and they want you to celebrate with them. Head to Branch IO about and sign their virtual guest book. You'll be entered to win one of a number of prizes including a $1,000 Southwest travel voucher, a personalized growth session with our consulting services team, or some sweet premium branch swag.
- [00:36]
Eric Sufert
Go check it out.
- [00:38]
Sponsor/Ad Reader
That's Branch IO about.
- [00:42]
Dan Taylor
The problem is that the distinction needs to be drawn between the competence of the economists and the correctness of their analysis.
- [00:53]
Eric Sufert
Welcome to the Mobile Dev Memo podcast. I am your host Eric Sufert and I'm joined today by Dan Taylor. Dan, welcome to the podcast.
- [01:00]
Dan Taylor
Great to be here, Eric.
- [01:01]
Eric Sufert
So we have a lot to talk about and I'm excited to have our conversation. Before we dive in, could you please introduce yourself to the audience?
- [01:09]
Dan Taylor
Sure. So I'm Dan Taylor and I'm a vice president in our global advertising business. I've been with Google for about 20 years, always on the business and advertising side. Today I focus on ads commercialization, focused on bringing our AI powered advertising products to market with a particular focus on measurement and our advertising platforms.
- [01:29]
Eric Sufert
Today oftentimes I'll have senior people from, you know, the bigger platforms on the podcast and you know, they'll have been with the company for like a pretty long tenure. And I'm just like, just always curious to ask like, what have. And it's maybe, you know, I'll lobby a softball here but like what's the evolution of Google been like over 20 years? Because I mean that's, that's a pretty remarkable amount of time to spend with a company. But, but also like the Internet has just fundamentally changed in that time. I'd just be curious to maybe just if you just riff for a minute about like what you've seen change in the Internet landscape and like, because obviously Google would be kind of a front row seat to that.
- [02:02]
Dan Taylor
Yeah, sure. I joined Google from traditional broadcast media and I was brought in as part of a crop of advertising executives to convince large brands and agencies that search engine marketing was important. So that was a moment in time. And about a year into Google we acquired YouTube and wanted to convince brand advertisers that online video was going to be important. And, and so that was kind of a moment where we were relatively early in the industry of consumers were spending a lot of time on digital, but the advertising landscape was still strongly and firmly rooted in linear television and broadcast media. And that was kind of a phase where we were bringing people along not only on the consumer shift, but the technology shifts. And then fast forward a few years to mobile. I think that really supercharged a lot of people's attention on how rapidly and quickly things were changing and how the tools and technology needed to adapt. And no one knows that better than you do. I think in terms of how you got started with mobile, dev memo and all of that, just consumer time being spent, the need for publishers, advertisers and agencies to adapt to those tools and technologies has really been how we've been spending our time the last several years have been following consumer trends, adapting our technologies, and getting advertisers and, and businesses to come along on that journey with us. In terms of like, the overall shifts, it's really just been about how are consumers spending time, how are marketers investing their dollars and measuring accordingly? I think most interestingly and more recently, I've been thinking about this shift from traditional media not being able to measure really that well at all to in digital. Oh, great, I can measure something. But we were largely measuring like proxy metrics to moving towards. Okay, I want to measure something that actually matters, whether that's profit or conversions to now, like, all right, I've largely got a lot of dollars spent in digital. I now need to understand which of those platforms and players are driving the most incremental return for my investment. So it's really kind of moving more and more towards business outcomes and not just proxy metrics today.
- [04:17]
Eric Sufert
Yeah, that's fascinating. So I mean, I guess kind of when you sort of survey your, your career at Google, I mean, I guess you've got like the consumer adoption of the Internet epoch, which like, seems like, I guess, quaint. Right. But that was a dramatic shift in behaviors at that time. And then the mobile epoch, I would imagine is it stands out right, as like a distinct period. And then now, you know, and this is kind of teeing up the first question here, but probably you see it as the AI epoch and how like all of consumer behavior and the way and advertising to a large degree is aligning around that. Is that. Would that be like a kind of accurate characterization?
- [04:55]
Dan Taylor
Yeah, I think so. I tend to look back, to look forward and I do see a lot of parallels to these big consumer and technology shifts. And there's A lot of parallels to the shift to consumers online, the shift to broadband and online video, the shift to mobile, and now today's shift with AI. Although the shift with AI feels faster and bigger than the ones that came before it.
- [05:20]
Eric Sufert
I spoke at IO, I want to say three years ago and one thing I was struck by was there was this policy track that I was on and this, I guess you would say at that moment in time we were sort of like in the early innings of kind of this AI consumer adoption cycle. Right. And the speakers from Google side, they did sort of highlight the application of LLMs across the consumer surface area within Google's portfolio historically. Right. And so, you know, I remember someone making the point and like I, I just never really thought about it, but it was like, what do you think autocomplete is? That's, that is we were just applying an LLM. Now it's obviously it's not comparable to like the scale of the models that we use today, but like they, they were kind of making that point that like, look, I mean this, this technology is new with respect to like a chatbot, but it's not new. And you know, obviously the attention is all you need. That was 2017, right. So like that's years and years. But that was pretty interesting. Maybe talk to me about that because like being inside the company, I mean, it would have seemed slower and more incremental evolution of this tech and then all of the sudden the consumers became aware of it. Right, but from your perspective, and it had just been something that it's been incrementally getting better and better over time. And now we've got something to release to the public and they'll be seeing it for the first time. But actually we've been using it internally for some time. Is that kind of how that evolved in some ways.
- [06:39]
Dan Taylor
And I think the, you're right, going back a really long way, thinking about how even when you go into google.com many years ago and you would start to type a sear query, you would get this autocomplete and a lot of that would be driven by what are other people searching for? And can I make that easier for you? Or smart reply in Gmail or autocorrect misspellings and you're starting to see that in workspace and things like that. And so I think we've started to see some incremental changes over time. It was this moment where consumer attention started to really go to oh, generative AI can really be seen and felt. But for probably a Good decade. We had been using predictive and analytical AI in our advertising and our consumer tools. I mean, go back to GDPR and att and we talked about advertiser metrics and measurement and optimization. We had built this promise up of being able to have a really strong way to measure advertising. And then that got fragmented really quickly as you went from desktop to mobile to connected TVs to privacy and browser changes. And we had to build a lot of technology to project and estimate conversions or audience behaviors where we couldn't actually see it. And that was all AI technology in the background. But as soon as you started to bring that stuff to the foreground, consumers started to say, oh, okay, this technology is real, but it's been real for a while. So I think maybe to answer your question in an end around way, I think the generative AI technology really just came to foreground in the last couple years in ways that consumers noticed.
- [08:22]
Eric Sufert
Right, right. And also it benefited from the rebrand, right? From ML to AI. Right. I mean, you know, oh, for sure.
- [08:29]
Dan Taylor
I think we were using the phrase automation and we're like, no, no, no, it's AI. Yeah.
- [08:35]
Eric Sufert
So one thing that the whole, you know, the sort of like prolonged ATT waiting period, a lesson that it bestowed upon me because remember, like they announced it and then they delayed it. Right? And so it was, that's why it was, it was supposed to be. I was 14, it was like 14.6. But one thing that taught me was, was how ill equipped a lot of advertisers are to adopt these technologies.
- [08:55]
Dan Taylor
Now.
- [08:55]
Eric Sufert
Maybe that's changed in five years, right? Because that was, that's some time ago, but like half a decade ago. But, but because I remember in that time, in that like waiting period, I got, you know, contacted by like large advertisers, hey, help us. We need to adapt to these conversion values, right? We need to build some sort of predictive mechanism for LTV against these conversion values. And it was like, okay, the very largest advertisers had trouble and maybe some could do it, some couldn't, right? And like, and so then you just, there you did see this dependency on the biggest platforms because it's like that's the only option for this happening is depending on, on the big platforms who have the tech, who have the expertise, right? Who have the ML engineers, the infrastructure to do this. It was really, that was a wake up call to me because I felt not, you know, I felt that, yeah, these tools are available and like, they've been democratized in a lot of ways by the open source packages that you could use. Right. TensorFlow. Right. For instance, you know, or Pytorch, whatever your flavor is. But no, even then it's hard. And that's why the largest platforms, I think then as a result of att, that's, I mean, PMAX probably accelerated as a result. Right. I mean, so that's why you've seen a lot of these systems develop. I think it is downstream of that now. It would have happened anyway. But I think there was an acceleration to that and it does show that capability divergence between the very largest platforms who have these massive armies of people that are very skilled, have a ton of domain expertise and advertisers. And that gap probably gets wider by the day.
- [10:16]
Dan Taylor
I'd love to touch on that one thing because it brings something to mind for me. There's actually two polarities that I've seen on that front. The first one is on the track that you mentioned, which is as there's been fragmentation in the identity and the tracking space. And we've built a lot of these AI powered systems to help advertisers kind of smooth out the bumps, to help reach audiences at scale and deliver on their advertising goals. The other thing that's happened though is that advertisers have really gotten smart about taking control of their own data assets, their own first party data. And so, you know, getting their own tagging infrastructure and analytics set up in house, starting to build their CRM system connectors. If you're a lead gen advertiser, really understanding what kind of data you can send to the platforms that you're working with to tell them which conversions you're sending them actually turned into a profitable sale versus which ones were useless to them in the end. And so two things happen, right. One is, yes, there's these AI powered tools that help them drive scale faster, but also they're learning that the data that I send into those platforms to drive results for me actually matters a heck of a lot more than just turning it all over to the platforms to figure it out. For me, I thought it was an interesting polarity that both of those things happen at the same time.
- [11:39]
Eric Sufert
That's a very astute point. Right. And that's something I believe as well. I mean, I've called this like the measurement renaissance.
- [11:45]
Dan Taylor
Oh, for sure. Like all of a sudden MMMs are like in vogue again, right?
- [11:48]
Eric Sufert
Exactly. You know, back to the talk about back to the future. So I had Carl Mela on the podcast a while back who's like, you Know, a legend in the brand marketing space. He's a professor at Duke. And we were talking about this and we were talking about like, you know, how funny is it that MMMs are like at the forefront of measurement now? Because if you told someone that in like 2018, they say, what are you talking about? Like, that's preposterous. Right. But it forced the discipline. It forced the discipline of relying on like kind of holistic measurement, sort of weaning off a lot of these deterministic signals because you lost access to them. And I do think in a lot of ways it enabled a lot, first of all, more rigorous measurement. Right. Which, you know, mmms and, but incrementality testing. And, you know, Google's got Meridian, but Google's led the way here for a while. They had the, you know, there was a Skunk Works project within Google that made lightweight mmm. Right. I mean, that wasn't even an official product. That was just a couple of people thought, people need to have this.
- [12:35]
Dan Taylor
Yeah. So again, like, that was a big investment for us and productizing that with Meridian last year was a really important thing for us. And not only open sourcing that, but, you know, working with third party partners to put those models to work inside of their MMMs to make sure that our media and the organic search priors and things like that are being baked in. So, yeah, it's been a really interesting again, back to the future moment, but also really critical in this AI powered moment because I think the data that you bring in to steer the AI actually matters a ton and it's a differentiation for marketers to get it right.
- [13:09]
Eric Sufert
Well, thank you for segueing me because I would have kept us here for the whole hour.
- [13:12]
Dan Taylor
Oh, good, I segued you. Yeah. Anyway, yeah, sure.
- [13:15]
Eric Sufert
Speaking of AI. Yeah. Okay, so Q4 earnings, Google reveals AI overviews queries doubled since launch. Talk to me about what's behind that. So are users adapting search habits to AI overviews? Are they writing queries that are more likely to sort of be answered or answerable by AI overviews, or has the AI overviews coverage increases? It's just more like a coverage effect.
- [13:37]
Dan Taylor
So people's expectations of behavior changed as we brought AI to search. And so it's a little bit of what we talked about a moment ago. They're seeing AI overviews, they're seeing AI mode, and it's like, okay, I can ask new and different types of questions, things I wouldn't have come to Google to ask before. Longer, more complex, even visual. We talk about Google lens, and that's 25 billion searches a month today. And so that's leading to growth in overall queries, including commercial ones. And so I talk about this thing where you no longer have to think about the right way to search on Google. You just ask the question. And so people are moving away from keyword ease to more conversational and intuitive experience. And that is what's triggering more AI overviews, and that's driving more usage of AI mode as part of that expansion. Along with that is coming Gemini's better understanding of these longer and more complex queries. So it's more about how user behavior is changing as opposed to how we're tweaking the experiences.
- [14:34]
Eric Sufert
It's funny that you mentioned that, because I was talking to someone the other day about how historically search had been a skill. Right. It was a consumer skill because you needed to know how to construct the query. Right. And, you know, to be concise, coax
- [14:47]
Dan Taylor
it out of the search engine.
- [14:48]
Eric Sufert
Yeah, exactly. And then, like, you know how to structure it such that, you know, you were putting the focus on the keyword that you cared about, but there was enough detail there to, like, maybe parse out different variations of the data. And, like, you just get. I feel like it's a more forgiving interface now, like when it's conversational. Is that. Do you think that's right? Is it?
- [15:06]
Dan Taylor
That's absolutely right. I mean, I was on a car drive last night and I used AI mode and just had a conversation. And it was easy. It was just easy.
- [15:16]
Eric Sufert
Right.
- [15:17]
Dan Taylor
And so it turns out when you make search easier, people search more.
- [15:20]
Eric Sufert
Right. And, you know, you just get more shots on goal in a sense. Right?
- [15:23]
Dan Taylor
Yeah, exactly. Yep.
- [15:27]
Sponsor/Ad Reader
Mobile game developers no longer need to pay up to 30% in major app store fees. With Xsolo Webshop, you can create a direct storefront, cut fees down to as low as 5%, and keep players engaged with bundles, rewards and analytics. Start today@xsola.com that's x s o l l a.com or use the link in the episode Show Notes.
- [15:55]
Eric Sufert
Okay, can you give any updates on AI Overviews, ads, monetization relative to legacy search? So the last data point was that AI Overviews is monetizing app parity. Is that. Is that still the case? Can we still say that?
- [16:06]
Dan Taylor
Yeah, that's right. So AI Overviews are still monetizing at Parity, which is a strong base. What's been really interesting on the ad side there, though, is with customers, is how AI Overviews is becoming A more powerful engine for brand discovery, creating new opportunities to get in front of customers earlier in the journey. And so what I mean by that, it's part of that expansionary moment that you hear us talking about. And so AI overviews, AI mode are matching ads not just to what the consumer is searching for, but also to the context of the answer. For example, when an AI overview is an AI generated snapshot in the results, the aha moment for advertisers has been an ad can appear not just based on what the person searched for, but what shows up in the AI overview. And so I can give you an example from my own life. So I travel a lot for business. My wife is now sometimes coming with me now that our son's off at college, we started talking about bringing our cat as the cat, I guess gets lonely. I don't, I don't really know, it's hard to tell. So we search how do I bring a cat on a flight? That's the type of search that triggers an AI overview. And our search models break down that question into smaller subtopics. Liz Reid calls that the query fan out technique. Right. And the AI overview shows airline policies you have to pay, apparently vaccine info depending on where you're going, how to keep them calm, and that you need a compliant pet carrier that all others up to that broader question. Now, in the old days of search engine marketing, most companies are not going to bid how do I bring a cat on a flight? But that pet carrier information in the AI overview triggers a new ad opportunity for companies like Petco, Walmart or a company called Cat in the Bag that sell pet carriers. And now that's a new ad opportunity that didn't exist before for advertisers. And so it creates new monetization opportunities.
- [17:57]
Eric Sufert
Oh, that's really fascinating. So you've got the sort of probably less distilled like from a keyword perspective query. Right. And it's almost like you're kind of parsing out the relevant keywords to then have bids be submitted against. Right. That's really fascinating. Is that kind of right?
- [18:17]
Dan Taylor
Yeah, I think the piece that was a little underappreciated by me at the beginning and many of our advertisers is that not only is it hard for me to predict what people are searching for, and my keyword strategy is not as nimble to these new ways that people are searching, but I also have these opportunities with AI overviews where what's in the response can be a commercial opportunity.
- [18:41]
Eric Sufert
Right.
- [18:42]
Dan Taylor
That wouldn't be I mean, you know, it brings people kind of down the funnel, if you will, as part of their discovery journey. And that's been part of the expansionary opportunity.
- [18:50]
Eric Sufert
Yeah, that's, that's really interesting. I guess I just never really considered it from that angle. But it, you know, it makes sense. There's just, it's. The stuff that you bring in itself has keywords, right. I mean, you've got this conversational query, this natural language query that's not keyword distilled because it doesn't need to be. And that's not how people talk.
- [19:06]
Dan Taylor
Right. But you break it down in your head. Oh, there's five or six questions inside of that question. Right, right, right. Like when you say, I want to plan a trip to San Francisco this summer, there's seven or eight things inside of that question. Right. Airline, hotel, things to do, weather, et cetera.
- [19:24]
Eric Sufert
Right, right, right. Well, that's a nice setup for my next question. So AI mode queries again revealed, this was disclosed in the last earnings. So they're 3x longer than legacy search. I mean, that makes sense, right? It's a conversational format, Right. It's conversational interface. How does that impact ad ranking? Right. So if we think of. There's a lot of additional context there. It's like we were talking about, like this was a skill before. I need to know how to prompt, in a sense or, sorry, know how to search. I need to know how to search such that I'm making a very, you know, dense nexus of keyword information that, that Google can then parse out to sort of rank the relevant links. Right now with AI mode, I'm just giving you a lot more stuff. And now just to be clear, I'm talking about now AI mode for the listeners, not AI overviews. So how does that. And so, and you know, this was touched upon in the Q4 earnings call. But I would just love to hear more about this. Like, how is this contributing to ad ranking? How is this giving more information to ranking those ads?
- [20:17]
Dan Taylor
Yeah, there's two pieces to that. So first, these longer, more conversational queries, they just give us more to work with, right? There's more context to better understand the commercial intent and a better signal for our ad systems to work with. Now historically, serving ads on these longer and more complex searches was challenging. Like that keyword ease thing we were talking about a minute ago. Gemini has dramatically improved our ability to better understand those longer and more conversational queries, particularly in non English language. We've really seen a huge improvement There that's helped us capture interest in new forms of search. The other half though, is our ability to predict which ads are going to perform best, which gets into the ad ranking question that you had. So for every ad, we generate a prediction of how well that ad is going to perform against a given query. Is the user going to click on it? Are they going to convert? In addition to better understanding the intent behind someone's query, we also have dramatically improved predicting how well an ad is going to perform against that query. We've been making these improvements to search query understanding at a rate of about one launch per month for the last two years. We touched on this on the earnings call too, leading to a 40% reduction in irrelevant ads, which for me I was like, well, our AD system for 25 years has been working quite well, but it turns out with Gemini, we found a whole bunch of new headroom on ads quality.
- [21:48]
Eric Sufert
I think this is like my favorite topic at the moment. It's LLM based ranking. But it's fascinating because to your point, I mean this is a domain that's had the benefit of 20 years of the best data scientists, ML researchers, ML engineers.
- [22:06]
Dan Taylor
Yeah. I mean, hundreds of experiments and changes every year. Yeah.
- [22:10]
Eric Sufert
And you're, you're seeing these, these significant improvements on that baseline and it's, it's fascinating. There's actually really, I don't have it pulled up, unfortunately. I mean, I'll link it in the show notes. There's a really great paper from Google on the use of LLMs for AD ranking, which it's because you just get so much more data to bring to bear.
- [22:27]
Dan Taylor
Yeah.
- [22:27]
Eric Sufert
In determining will this convert, like will the user click this, does this match what they want? And you get like, you know, this just semantic understanding to apply to that. Whatever the input is, whether it's just a, you know, they're browsing YouTube or whatever or they're searching or whatever, they're on a social media feed. Like there's some context and it's very shocking to me, like how big the improvements are here.
- [22:53]
Dan Taylor
Yeah. I mean it turns out that, you know, prediction is the perfect problem to put AI against.
- [22:58]
Eric Sufert
Right. But talk to me on that topic, that expansion, I want to kind of anchor to this idea and I think this is a couple years old, so I don't know if it's like it's current, but it's kind of an infamous statistic that Google shared is that 80% of Google search queries were not monetizable.
- [23:13]
Dan Taylor
Right.
- [23:13]
Eric Sufert
There's just no commercial intent from this skill that the user has needed to enable to get the right links out of the search result. How does that change in the AI overviews? I think you kind of hinted at this, but I'll just sort of ask it more pointedly in the AI overviews. AI mode in interaction models, is that part of the expansion here? It's just like, well, anything can really be, you know, sort of commercially tenable because we can parse out a lot more information in this conversational format.
- [23:39]
Dan Taylor
I don't think we've updated that stat, so I don't know if it's a current one or not. But our core philosophy remains the same, right? We show ads when there's commercial intent. What has changed is our ability to spot it. And so you touched on that. It's true that the vast majority of searches never involve an ad because ads are meant to be helpful to your search. And so if you're looking for today's weather, the most helpful thing is going to be a direct answer, probably not an advert. We're seeing more than 5 trillion searches on Google a year, but the overall number of queries, including commercial ones, are going up. That's largely because of our investments in AI like AI overviews. We talked about ads above, below and within AI overviews and we're feeling good about the results we're seeing there, including new opportunities as we think about these new experiences in AI mode. For example, we're just testing ads in the US today, mostly focused on getting the experience, the user experience, right. AI mode is something between a chatbot and a traditional search engine. We're reimagining search with AI, a conversational experience, but grounded in the full information of the web. And so in our test, what I think is really interesting to touch on here is we're being thoughtful about where and how an ad might show in a person's journey. So we've got, you know, we're testing some formats and things like that. But a key learning from our testing is not where on the page, but where in the journey an ad is served. Right? This isn't like enter a query, get a result, move on. It's about when ads are useful and relevant, they're appreciated, they don't disturb the experience. And so if it's a conversation and a multi query journey or session, if you will, putting an ad too early, not going to be a good experience. So an example, about a year ago I got into running, I'm ready to move to 10k and I don't run that far on a regular basis. So I'm going to use AI mode for tips on how to train for longer distances. What's a good heart rate, variable weather this time of year on the east coast, nutrition, things like that. It's a longer, more complex query. I'll probably do a couple of follow up questions on informational needs, but if you throw me an ad for new shoes right up front, you've kind of lost my trust. Not really a great experience. But if I dig in and ask follow up questions I could learn that like collagen could help my joints because I'm an older guy. There's some opportunities where ads might be welcome and the right answer to my need. And so we're really figuring out where in the monetization, we're where in the experience monetization makes the most sense in things like AI mode. And so that's what we're looking at right now.
- [26:05]
Eric Sufert
So maybe just, just to, just to dig in here for a bit. Yeah, sounds like we're talking about is kind of stacking the, the conversation based on I call it like the level of insight delivered. Right. So if I'm just kind of, kind of high level conceptual question ad wouldn't make sense. It's too aggressive. But like as you sort of like deliver more and more information to the person to such that you know they're now they're like kind of more primed to buy then that's the right time to show an ad. Or are you thinking about. So I guess the question is like are you thinking about this within a single conversation or is this like across multiple conversations over time is the question.
- [26:38]
Dan Taylor
I guess I break that into a few things with AI overviews. Taking that back, that's an extension of more or less traditional search, right? So it's a, it's a representation of the traditional search result that has more information and when that's an expansionary opportunity and we're finding new opportunities to monetize there both by understanding the longer queries better and also the context of the page which we talked about in AI mode, we're finding opportunities to monetize in those longer, more conversational back and forth. So we're trying to find the right user experience and the right ad formats to introduce there. I'm mainly talking about at this moment like within a single conversation, but you could see how that might be something you would introduce over time. So you know, one of the things that consumers expect from their AI powered experiences is contacts. They want to make sure that they understand Experiences over time.
- [27:30]
Eric Sufert
Right, got it. Okay. So I wrote a multi part series, also did like a podcast about it called Google's Gambit. The idea of Google's Gambit was that, you know, there's this absorption of engagement that used to be directed externally. Right. As a result of the links into AI overviews and AI mode. So in that, you know, in that piece I wasn't making a normative argument, it was just an observation of incentives. Right, that makes sense. Right. Especially given the consumer shift into AI interactions. So what do you think? I just, I would just love to get your like kind of big picture sense here. Like what are the long term consequences of this on Google's advertising business? Where and where does this take search?
- [28:04]
Dan Taylor
So we do see this as an expansionary moment. Overall I think AI is helping people ask new questions, enable businesses to grow in new ways. But we also see it creating new opportunities for creators on the web. We talked a little bit earlier about this is a big technology shift. This reminds me that pivot to mobile. We talked about changing how consumers are engaging with content and platforms, including our own. As we're enabling more and different types of question with search and AI, our ads focus is creating opportunities for businesses, more relevant ads, but also providing improved tools for content creations. You know, whether that's Asset Studio and ads or all the tools that we're giving to YouTube content creators as well. I think as consumer behavior and time spent on these new experiences evolves, we'll experiment with formats, with measurement to adapt to those business needs. But I also see AI enabling greater discovery on the web. Right. Helping people go deeper in the research, discover sites and new brands and authoritative sources they might not have otherwise found. I think that is one of the things I personally have enjoyed and one of the things we're focused on in the Google search experience in particular. It's grounded in the information around the web. And so one of the things that is really interesting about AI overviews is it very much focuses on here are the authoritative sources around the web. Here are qliks, go link and learn more. And perhaps more than any other company, we're committed to making sure that we're sending quality traffic out to the web. And I think that's remains central to our approach.
- [29:34]
Sponsor/Ad Reader
You know those channels your colleagues keep bragging about? The ones getting all the credit? Yeah, they might be doing squat attribution makes every channel look like a hero even when it's a zero incremental, tells you who's actually doing the work. It's like A lie detector for your marketing budget. Start using incremental today. Get your demo@ incrementl.com that's incrmintal. Mention that you came through the mobile dev memo podcast for a special 15% discount for the first six months.
- [30:08]
Eric Sufert
Can you talk to me about how Gemini is being applied to the ad stacks that was discussed in the earnings call.
- [30:12]
Dan Taylor
Yeah.
- [30:13]
Eric Sufert
And you know, ranking rankings, one piece of that. Right. What are the other touch points? Right. And I'd actually love to hear anything that you could share about, like how it's being applied in the YouTube context. But you know, we've got obvious applications like creative Generation. Right. That makes sense. I think everyone understands that. We've got LLMs for AD ranking, re ranking that we talked about. But like talk about intent classifications. What are the applications that maybe surprised you too? I'd be interested to hear that.
- [30:36]
Dan Taylor
Yeah, for sure. So we're deploying Gemini models across Alphabet, which means that when Gemini gets an update, ads gets an update. So Gemini 3 in our search models or nano banana 2 in Google asset Studio. I start with that because while that may seem obvious, but given the pace of AI innovation and launches, that speed to market has been a real advantage for our customers. I think about Gemini and our ad stack in three ways. Number one, new AI user experiences like AI overviews and AI mode. It's inspiring greater search usage. Number two, we're better understanding the new ways that people are using search, these longer and complex queries. And we're increasing our ad quality based on that. We talked about that too. Third of all, which we haven't talked about a ton yet, advertiser tools are making it easier to capture these new opportunities. And so I'll talk about that a little bit. Three things on advertiser tools, campaigns, creative assets and in product advisors. I think on a campaign side, the days of kind of keyword lists are fading. We talked the keyword ease thing. Consumers are no longer really searching in fragments, they're searching conversationally. And so our AI powered campaigns are helping partners adapt to this new way of searching. And so AI Max is a good example of this. Performance Max is another. It's helping advertisers who have search campaigns or keyword based campaigns expand in these new ways that people are searching and reach billions of net new searches. They simply weren't reaching before. On the creative side, whether it's tools like Nanobanano and VO3, Gemini's fueling creativity at a scale that would have been really impossible due to time and budget constraints, even a Couple years ago. This gets deployed by advertisers in AI Max and Performance Max and also in Asset Studio and then finally in Product Advisors. These are things like Ads advisors in Google Ads and Display and Video360 or Analytics Advisor. These are kind of like in product assistance that help you set up campaigns, find new insights, troubleshoot ad creative that isn't running and things like that. Just less time hands on keyboard and more time uncovering insights and moving faster. So all of those things really help advertisers capture the opportunity for the new ways that people are searching and as I mentioned before, take advantage of these new updates as soon as they ship.
- [32:58]
Eric Sufert
I'd love to just hover on the campaign management piece for a second because I think that's one thing that's underappreciated is how long that's existed within the Google ecosystem. It had a different name. It was uac. Yeah, Universal App Campaigns.
- [33:16]
Dan Taylor
I was on that project back in the day.
- [33:19]
Eric Sufert
Well that was the OG pmax. I mean that was the original, that was the original end to end automation system. And I wrote a piece about it when it was first introduced which was the idea was like hey, we're going to consolidate all these app campaign. So just to clarify for people that are not familiar UAC and you could search on mobile dev memo for I don't know what it was called but Universal App Campaign, search for that, you'll find it. But this was 1718 I think, do you remember earlier? Earlier 16, yeah. So the idea was like hey, we're going to consolidate all these disparate campaigns that you've got running into this one system and we're going to manage a lot of the variables that you normally would attend to in an automated way. And you know, I wrote a piece that somehow managed to upset Google and upset all the readers which was that look, you know, this helps Google, this is good for Google. But hey, advertisers, it's good for you too. Because you know what, you're probably not as good as Google as this and you know that. I think that argument is still valid. I wrote a piece a while back called Satisficers Remorse which basically makes the same point, like if they're meeting your needs, you're better off. And so even the control maybe has been surrendered to some extent but like you're still better off and you probably couldn't achieve the same outcome on your own. Right, but that was the point I was making with UAC as well. But like that worked really well and So I pointed to that, I think in the Google's Gambit podcast. But maybe it's a different podcast, but I made the point that, look, if you looked at that was the original one UAC and then a meta followed or Facebook at the time followed with like AEO and VO and stuff. And as those tools got brought to market, ad spend went up. There wasn't this mass extinction event of UA managers. In fact, there were more hired. And I even had a bit in the podcast about hiring trends using, I think, BLS data. So it's like the market expanded. This was expansionary, and it was because. And the biggest platforms adopted this. And so I was making the case that there's this panic about AI and mass layoffs. But if you look back at the history of what these things have done, they've been very expansionary for digital advertising. And you could make the argument, I made the argument for the broader economy as a result.
- [35:37]
Dan Taylor
And what else went up was app attribution providers to help make sense of how well it was performing and how you should report back on the measurement aspect to steer the AI in the right direction. So it's back to that earlier conversation. The data that you provide, those tools and systems to steer it in the right direction actually matters, right?
- [35:54]
Eric Sufert
Well, yeah, because that birth, these systems birthed what's now known as signal engineering, but back then would have been just like figuring out like, what's the right postback, what's the right postback to send to indicate that this is a quality user. That mean that kind of birthed that whole feedback loop.
- [36:10]
Dan Taylor
Yep.
- [36:11]
Eric Sufert
And it's, it's like, you know, in kind of going to the very start of the conversation, it's like, look, I mean, I get that it's maybe uncomfortable for advertisers, but the reality is, you know, you look at the investments being made by these huge platforms, you can't match that. You can't match the domain expertise. You don't have these kind of, this kind of talent internally. You just can't. It's not possible. And so it's like, yeah, there's incentives involved, of course, we're in an economy. But like, at the end of the day, you're getting access to firepower that you wouldn't have access to otherwise. And if the outcomes meet your expectations, which is which, by the way, is the entire point of these systems, because that's what you provide it with, that's the input. There's nothing really to complain about.
- [36:54]
Dan Taylor
Well, look, 12 of my 20 years at Google have been in like the product go to market side of the house and I spent a lot of it on audiences and inventory and that side of the house. But you know, the last few years I've been really focused on data and measurement because in an AI powered world, to your point, like if that's steering a lot of the decision making on an individual ad placement, where the leverage and where the fulcrum is, is what outcome you're telling it to optimize for.
- [37:29]
Eric Sufert
Right.
- [37:29]
Dan Taylor
And where I see most marketing organizations pivoting their energy these days is there. And so I think that is the right move.
- [37:40]
Eric Sufert
Yeah. And that's, that's, you know, what I've called or I'm not the one that came up with the term, but signal engineering. Right, that's that whole exercise.
- [37:46]
Dan Taylor
Yeah, I like that term. I actually hadn't heard that one. Yeah, yeah.
- [37:49]
Eric Sufert
So I picked that up from the CEO of Voyantes. I'm an investor in Voyantis, by the way. Disclosing, but I think he came up with it. But it's just the discip of recognizing that the onus is on me, the advertiser, to determine how good is this user.
- [38:02]
Dan Taylor
Yeah, yeah. AI is only as good as the fuel you give it.
- [38:04]
Eric Sufert
Right, right, exactly. Okay, I want to switch gears a little bit. Can you talk to me about direct offers? Talk to me about, you know that, that launched what, two months ago? Tell me how it's going.
- [38:13]
Dan Taylor
Yeah, we announced it in January at nrf and so it's a good example of how we're looking to reinvent ads for the new era of search. So it was a new type of ad format that we're in piloting, built specifically for AI mode. Since we've only kicked it off at the beginning of this year, I can touch briefly maybe on how it works and then what we're seeing. So I'll start with an example. It basically introduces an offer in the AI mode experience based on where the user is in the journey. So it talks a little bit about how we talked about where someone is in the journey as opposed to just showing them an ad immediately. So I recently got a new laptop. It's bigger than my old one, so I need a new bag. I used AI mode to research stylish neoprene backpacks suitable for men in business settings that are under 200 bucks can fit a 14 inch laptop. So I asked Google a couple refining questions. How many pockets does it have? And then Google elevates the most relevant products to meet my needs. But often you're only ready to buy if you're getting a great deal. And so with direct offers, relevant retailers give a special discount like 20% off or free shipping, which is what I'm looking for. And a sponsored offer can be helpful at just the right moment, helping you get better value while assisting the retailer in closing the sale. So it's still early, but we're seeing some good product market fit and really strong interest from advertisers. So we're testing with Petco, Elf Cosmetics, Samsonite Rugs usa, Chewy, l' Oreal, I'm and Shopify merchants. We're hoping to have more to share in the coming weeks. I don't have any performance stats to show, but it's promising. Experiment. I think it's a good example of the types of things that we're looking to experiment with where in a traditional search engine marketing campaign you would just show that offer to everyone. But in an AI mode experience, we want to be able to understand where someone is in their journey and find, oh, this is a moment where someone's ready to buy and this could be the thing that gets them over the edge.
- [40:09]
Eric Sufert
Got it. I think it was at the same event, was it the Retail something Forum, National Retail Federation. That's right. Okay.
- [40:18]
Dan Taylor
Oddly held on a Sunday, but that's right.
- [40:20]
Eric Sufert
I remember being bewildered like how's their news today?
- [40:23]
Dan Taylor
But I guess because you got to get back to work. Yeah.
- [40:27]
Eric Sufert
So I think also announced at this event was ucp, is that right? Or is just thereafter if it wasn't at the same event?
- [40:33]
Dan Taylor
Yeah, that's right, yeah. Universal Commerce protocol.
- [40:36]
Eric Sufert
Okay. Maybe just kind of give us the same style rundown. Like how has that been received? Like how is it working?
- [40:42]
Dan Taylor
So I think at a high level agentic or AI assisted commerce is starting to become a reality. We've had a lot of interest in response to that Universal Commerce protocol. And I think the reason for that, it's a common language to better enable businesses to connect with AI agents as part of the shopping journey. Right. So we've been focused on making sure that we get the building blocks in place to support the shopping journey in this new era. From discovery to checkout. So we built this in partnership with retailers, but it's an open source protocol so anyone can use it. We're of course using it for new shopping experiences on Google. So we started by rolling out a new UCP powered checkout so in the US shoppers can buy items from companies like Etsy and from Wayfair. Right. In AI mode, in search and also in the Gemini app without ever leaving the conversation. And so at Shop Talk actually in March, we announced new capabilities like adding multiple items to the cart catalog features and supporting loyalty programs to make shopping more connected across the web. And companies like Commerce Inc. Salesforce and Stripe also announced they're going to implement UCP on their platforms in the near future. So I think at a macro level, the industry is still in pretty early stages, but are kind of flocking to this notion of commerce protocols for the technology to scale so that buying agents and retail agents can talk to each other in a way that is frictionless and secure and that retailers can still own that transaction and that relationship with the end user needs to be open and collaborative. And by all accounts that seems to be how this is being received. So it's been really successful so far.
- [42:28]
Eric Sufert
I think one of the questions I had when I first read about this was like, what's the tie in with ads? I mean, is this specifically for ad interactions or is this for anything?
- [42:37]
Dan Taylor
There's not a tie in with ads directly. I think that it's just really important for us to make sure that we enable retailers and consumers to have a frictionless experience. I think that consumers love the fun parts of shopping like being inspired by a new brand or a virtual try on and things like that. They don't love the friction filled parts like find it at this size or find it in this price or fill out my CVV code and remember my loyalty program. And so these sorts of protocols take the, the friction out of the shopping and that's the piece that we're focused on.
- [43:11]
Eric Sufert
So I remember my reaction and you know that's clarifying, right, that it's not distinctly as product. But my reaction at the time is so if you imagine like this kind of protocol space is going to get pretty fragmented and it already is, right, to a certain extent, but it probably become more. So the reason someone would be incentivized to adopt UCP is because they're already buying ads on Google, right? That's got to be a motivation for adoption to some degree, right? I'm already buying ad. If I'm going to prioritize these, what might I prioritize? Very well. That's probably the place where I'm buying ads and then I can integrate in that way. I mean, does that make sense or am I off?
- [43:48]
Dan Taylor
We're thinking more expansively than that. So we've got a shopping graph of 50 billion products. We want consumers to discover products across on Google whether they're buying ads from us or not. And so there's 50 billion products that are discoverable on Google and we refresh 2 billion of them every hour. And there are many merchants that buy ads on Google to support that, and there are many merchants that don't.
- [44:13]
Eric Sufert
All right, I'm cognizant of the time. I want to kind of just finish on just one question. So what do people most commonly get wrong about Google's advertising business? What's the most common misconception? What's the sort of like highest density misconception you could disabuse somebody of? What are people just generally wrong about when they think of Google Ads?
- [44:32]
Dan Taylor
A common misperception is that search is purely a lower funnel tool. Google search doesn't just close the loop for brands, it often opens the loop. The reality is that search is a massive engine for brand discovery. Over 70% of shoppers come to Google search open to trying new brands or products. And that's actually expanding with AI. And we talked about that a bit. And of course YouTube offers incredible opportunities for brands to tap into moments of discovery. The connection between the two is also interesting. People come to Google for everything from quick questions to high stakes decisions to engaging with trusted creators. And we're helping bring more of these insights to advertisers with tools like attributed. Branded searches is just one example where we can report to advertisers when consumers search for a brand after they see a video ad on YouTube and ways to engage differently with them based on that, we'll have more on that in the works. But we're really focusing in on helping advertisers understand how they can capture opportunities with consumers when they're earlier in their decision journey. Because there's a tremendous amount of opportunity right within Google. Search is not a place where you just capture demand. It's where discovery starts and decisions are made.
- [45:46]
Eric Sufert
Dan Taylor, I appreciate your time. I appreciate you chatting with me. Thank you very much. And I appreciate you and the whole team for making this happen. Thank you very much.
- [45:53]
Dan Taylor
Yeah, Eric, it's a pleasure. Thanks so much.