← Next in Media
Next in Media
How Google Reinvented Search with AI
00:27:33
Google Search just turned 25, and it’s entering a new era—one driven by AI, conversations, and context. In this episode, Dan Taylor, VP of Global Ads at Google, joins Mike Shields to unpack how the search giant is transforming its ads business for the age of AI Overviews, Performance Max, and long-form conversational queries. He explains why this shift feels bigger than mobile, how advertisers are adopting AI faster than ever, and why trust and accuracy remain Google’s north stars. Dan also reveals how AI is expanding the search funnel, creating new commercial moments that brands never could have targeted before. From tools like AI Max to agent-powered shopping, the future of advertising is about reducing friction, improving relevance, and meeting consumers wherever their curiosity starts. It’s a rare inside look at how Google plans to keep Search indispensable for the next 25 years.
Loading summary
Transcript76 lines
- [00:00]
A
I think today's shift with AI feels bigger, but it is oddly familiar at the same time in terms of a massive change in consumer behavior and marketers kind of working hard to keep up with the speed of the consumer.
- [00:14]
B
So do you see this as like, as profound as we make it out to be?
- [00:17]
A
You know, it takes a long time to build trust and only a moment to lose it. Right. And so because people rely on the information they get from Google, we really wanted to have a high bar for quality and accuracy.
- [00:26]
B
How it is changing and will change advertising.
- [00:29]
A
80% of Google Ads advertisements are already using at one AI powered thing. And you know, 2 million of our advertisers are already using generative AI creative tools. We had a pretty big program called enhanced campaigns around 2013 that helped advertisers figure out how to apply their buying strategies from desktop into mobile and then tablet. One of the most successful launches I've ever been a part of in the ads business.
- [00:55]
B
This week on Next to Media, I spoke with Dan Taylor, VP of Global Ads at Google, upon the 25th anniversary of the company's storied search business, a business that is still dominant for now, but facing massive changes in consumer behavior and renewed competition. Dan and I talk about whether Google is moving fast enough to take on the lights of ChatGPT and perplexity and whether search ads will remain as effective in complex conversational search queries as they've been over the past two and a half decades. Let's get started. Hi everybody. Welcome to Next to Media. I'm Mike Shields. I'm here with Dan Taylor. He's a vice president at Global Ads at Google. Hey Dan, thanks for being here.
- [01:30]
A
Hey Mike, good to see you.
- [01:31]
B
Lots to talk about upon what happens to. We're talking two decades of Google search, which is pretty amazing.
- [01:37]
A
Yeah, yeah, 25 of Google search.
- [01:39]
B
I'm sorry, 25. 25. Before we get into like that and that like the state of the very interesting time we are and find ourselves in a search, just give me a little bit about your history there. Like tell me about your Google career. What was your first project?
- [01:51]
A
Yeah, yeah, sure. So I joined Google in 2006. So I've been here almost 20 years. I came from a broadcast advertising sales background. So you know, television, radio, traditional media as they call it. And obviously I was drawn by the consumer shift to people spending more time online, but also this, this notion of, you know, being able to measure advertising and adjust and adapt on what's working in closer to real time. So that was pretty Attractive to me.
- [02:21]
B
Even though that was probably like a killer job, right. Broadcast sales at that time. It's probably scary to make that move.
- [02:25]
A
It was, I mean it was one of those, like, I put my pinstripe suits in the closet, go find myself a pair of jeans and come to the office. It did. I had spent a lot of time obviously building relationships in the national sales arena and working with ad agencies. But interestingly enough, my first project was kind of a bridge. And so the first project I worked on was bringing digital ads like buying capabilities into traditional formats like radio and television. And so the project was ahead of its time in 2006 in terms of can we apply addressable advertising technology to traditional broadcast and traffic systems and measurement?
- [03:09]
B
So this was Google TV ads and then there was a radio company you guys acquired, if I remember correctly, around then.
- [03:14]
A
Yes, Yep, yep. DMARC was the name of the company and Google Audio Ads was the, was the offering. And so we learned a lot. We made some inroads in that, in that space, but definitely was ahead of its time. So worked on that for a couple of years and then moved into the search and YouTube sales display and video programmatic, kind of kept moving to the next emerging space, mobile and so on. And the current role that I have is helping connect advertiser needs with our product roadmap. And so it's a strategy role across all of our products.
- [03:49]
B
Yeah, that's funny because TV is doing all that stuff. It just took a little longer than everybody maybe expected or wanted at the time.
- [03:57]
A
Yeah, for sure. Like you know, connected TV and podcasting, like, you know, and all these new environments. So those early investments did pay off as we started to think about how to not only build partnerships and relationships, but how to monetize and understand what advertisers and publishers in that space care about. Right.
- [04:14]
B
Okay, so let's talk about this like fair. It feels like a fairly dramatic moment in the Internet and in digital advertising and in search. The shift towards conversational answer driven search. It's funny because some people talk as if it is the. A whole new Internet. Everything we know is changing and that other. Others are kind of maybe hold on a second. This is like this zero click search world. We've been getting there gradually over time. So I guess do you see this as like, as profound as we make it out to be or is this like an evolution or culmination of something?
- [04:46]
A
Yeah, I think profound is a good word, but not an overnight sensation, if you will. Right. So I think, you know, on top of many years of innovation, which is in our DNA. People are fundamentally changing how they search, how they discover information, how they interact with technology. And that creates new opportunities for advertisers to plug into those moments and be there in the relevant moments for a consumer with a product or service. But I was reflecting back on the 25 years and Sundar, when he first became CEO of Google, you know, one of the first things that he said was that we're going to be an AI first company. And that was 2016. And so that I don't even think.
- [05:28]
B
I, I don't even think that registered with me as much back then as it does now.
- [05:32]
A
Yeah. And so that's 10 years ago. Right. And so that started to manifest, at least in the ads business with more predictive and analytical AI, helped inform things like smart bidding and search modeling, conversions when we couldn't necessarily measure them directly. And some of these AI powered advertising tools like Performance Max and so on, and so the generative AI piece and what consumers experiences see that are different, that feels much more immediate, like wow, like AI is suddenly here. But you know, we've laid a lot of groundwork and so it's helped create an expansionary moment for search and that does feel big and it does feel profound.
- [06:14]
B
I want to get into how it is changing and will change advertising. This is compare it all to you. Search was still search on mobile, but it was a big moment at that time where people are like, oh, this is not going to maybe, maybe perform or behave as the same way as it had. Does this echo that period at all for you?
- [06:31]
A
Yeah, I think today's shift with AI feels bigger, but it is oddly familiar at the same time in terms of a massive change in consumer behavior and marketers kind of working hard to keep up with the speed of the consumer. And so I remember, you know, initially talking to customers about they should be advertising in mobile and it was, it was a hard sell. Right.
- [07:00]
B
It's a small screen and what are people going to do there?
- [07:03]
A
Like screens are small, the ad formats weren't fit for purpose, the buying tools definitely didn't work, the measurement didn't work. And so, but you know, after that initial period of kind of ignoring the problem and hoping it would go away or copy and pasting the same stuff, that's always my strategy on desktop. Right. Duck and cover. But then, you know, tools like how do you build a mobile website? How do you embrace the app economy? How can you adapt our tools? We had a pretty big program called enhanced campaigns around 2013 that helped advertisers figure out how to apply their buying strategies from desktop into mobile and then tablet. We also had a separate tablet strategy at the time. Right. That was kind of emerging. And so once kind of, we built some tools around, you know, bidding differently on mobile, new ad formats, cross device measurement. It unlocked a ton of success. I think it feels a little bit like that moment. But what's different with AI is that advertisers are embracing this shift much faster. In part because we've been on this journey for over a decade with some of the things I told you around getting customers to adopt these assets.
- [08:19]
B
Maybe they waited last time, it wasn't good. Let's move fast now.
- [08:22]
A
Yeah, I thought one of the things that was really interesting, I get this question a lot is how do you get advertisers to adopt these new AI powered experiences like AI overviews and generative AI creatives and things like that. In many ways we've been laying the foundation for that. And so bidding, reaching audiences, asset driven creative and these fully AI powered campaign types that we've had, that allows you to be much more nimble when these new surfaces come up. And so 80% of search, excuse me, of Google Ads, advertisers are already using at least one AI powered thing. And 2 million of our advertisers are already using generative AI creative tools. And so it was one of those things where we weren't, it wasn't a cold start in terms of getting advertisers into these new experiences.
- [09:13]
B
Speaking of cold start, like, I think there's a, there is a perception that, or there's a narrative that given their explosion of ChatGPT and others, did Google move a little too slow on AI? Like what do you, what do you say to that? People ask that question.
- [09:28]
A
Well, first I refer back to what we just talked about, which is we've been on a AI journey for, for 10 years and you know, core to everything we're, we do, from search to maps to ads. Right. You know, you think about Google Maps being to tell you how to take a different driving around traffic, right? Yeah, AI is in there. But I think the generative AI piece in particular, we see that it has the potential to change the world. We're excited by that. We put out AI principles again, I was reflecting back in 2018 to take this approach that we want to be bold, but we also want to be responsible. With 5 trillion searches every year, this requires us to be helpful and get it right. You know, it takes, it takes A long time to build trust and only a moment to lose it. Right. And so because people rely on the information they get from Google, we really wanted to have a high bar for quality and accuracy. And I think, and again, being in ads, you know, we see time and again if you just focus on building a good user experience, building a successful business around, it follows I tell my teams like, hey, it's not a race. We weren't the first search engine, we weren't the first email service, we weren't the first map or video website. And if you focus on building a great experience and being helpful, the business comes. And so we're already seeing that with like AI overviews which are being used today by 2 billion people, and Google Labs, which is like this area that we've built within Google to launch new and experimental features that allows us to kind of experiment at scale without moving it into the core experience until we feel like it's, it's, it's right.
- [11:11]
B
One of the world's most prolific creators and distributors of premium content such as America's Got Talent, the Price Is Right and Family feud, reaching over 100 million US households, leverages Elemental TV's AI powered one audience platform to offer advertisers access to its extensive library of iconic titles, all through a leaner optimized supply path. Lesser intermediaries enabling increased brand safety and more cost efficient campaigns. Discover hal@elementaltv.com One audience that's E L E M E N T a l t v.com One audience thank you. I want to maybe come back to Labs in a moment, but getting back to like your idea that if you do it right, the business will come. I think that initially when, when AI overviews and conversational search hit, there was a. I remember a few people writing, you know, I don't know if advertising is going to work here like this. This might just be not the right, the right canvases is not what people are asking for. Will it be welcome? What have you found early on? Like, is it just a take a different approach? Is it as welcome as a traditional search?
- [12:16]
A
We're not assuming it's going to be all the same plays will work in these new environments. And so like for example, the average length of a search query in AI mode is like two to three times longer than what you'd see in a traditional query.
- [12:30]
B
And so that's really different.
- [12:32]
A
Yeah. And so yes, AI helps us better understand those longer and more complex queries and what question that we're asking. You know, a lot of this Comes down to when is that right moment to be helpful? I think one of the things that I've loved about being at Google in the ads business is this core value that, you know, ads need to be relevant to the consumer, needs to be additive to the experience. And our business model rewards this. Right. You know, we don't get paid in search unless a consumer clicks on an ad or, you know, even in YouTube, like skippable TrueView ads. I remember when we first came out with those, like, it was one of those things where if your ad's not relevant, what are you doing enough for users to engage with it? We don't make any money. The advertiser doesn't succeed. And so it's a little bit.
- [13:17]
B
That's a new playbook for sure. Yeah.
- [13:19]
A
And so we're thinking that same way, obviously, in places like AI mode. I think what we've learned in AI overviews, we're starting to apply some of those things into our experiments with AI overviews. And so our key learning here is where in the journey, not the size of the ad or the format is. It's really the most important factor for feedback. And so do. If someone wants to talk about training for a marathon, you know, that, that could come to us as like a beginner type query. Where putting in, putting running shoes in the first, you know, first result is probably a little early. Right.
- [14:01]
B
Where, where the old search. If you just said what kind of shoes should you wear for the marathon? Or whatever that's going to be, that's you're. You're probably at the end of that.
- [14:08]
A
Journey, you're closer to, well, almost a little different. Right. So like how to train for a marathon doesn't even say anything about I need shoes. Right?
- [14:14]
B
Yeah, Right, right, right.
- [14:15]
A
And so that's going to get, that's going to generate like foundational advice, like a training regimen and nutrition and, and you know, maybe presenting ads for running shoes up front is going to be a miss. But alternatively, if there's an AI mode query about prepping for the New York Marathon, which is, I think two weeks from now, that person's going to likely need really specific information like whether what clothing to wear, you know, whether you need extra electrolyte supplements because it's going to be a hot day, like, those are the places where, okay, that this may make sense to show an ad in this moment. So that's the main thing that we've learned is that when someone enters AI mode and AI overviews, like an ad might not be the first thing that they're looking for, or actually let me rephrase that. A product or service might not be the thing that they're looking for. Right. And so, but because of the way that consumers are interacting with AI overviews and AI Mode, there are opportunities to introduce the ad. And that's the main thing that we've been focusing on is when is it additive, when is it helpful? Because it can be pretty off putting if you're going for information and a bunch of products get thrown at you in this experience.
- [15:25]
B
On the flip side, you have, you've talked about this, I think where this, the, the nature of longer queries and conversational search, it can open up new commercial moments that maybe weren't either readily available through the traditional search or they were different parts of the funnel. Can you talk about that?
- [15:46]
A
Yeah, for sure. And in fact, I think this is the aha moment for, for many of our advertisers. You know, AI in search has been expansionary for, for Google. People are coming to Google earlier in their discovery journey than ever. They're asking kind of these broader questions. They're not, they're not searching like a search engine marketer would with like very specific things that they're looking for. And this is a new opportunity to introduce consumers to products and services. They may not even know that they should be asking for this type of thing. And tools like AI Max were specifically built for this use. Case in mind, where you can't necessarily predict what these longer, more exploratory searches are. And I'll give you a personal example. My son is in his second year of college. He's, he's upgraded his dorm room. It's now kind of a apartment like thing with a kitchenette. Yeah, I know. If, if only I had as nice a dorm back in the day. He's got views of the city anyway.
- [16:48]
B
Did I ever cook?
- [16:52]
A
There's a kitchenette, but there's no place to sit and eat. And so space was limited. And we're like, let's, okay, let's let dad show you AI Mode here about how to furnish a tiny apartment. And we didn't really know what kind of things were available there. We just knew we needed something. And AI Mode used Gemini in this thing called Query Fan out, which is kind of a, you know, engineering term. But basically it means how do you take one long query and turn that into several shorter queries to come up with a lot of information that could be relevant. And so in a long query there's.
- [17:26]
B
A bunch of different pieces that could lead you to in different directions, is it?
- [17:30]
A
Exactly. And so the, you know, Gemini kind of breaks it down into like, well, like, this person could be looking for, you know, space saving. This person could be looking for, you know, small furniture. This person could be looking for cooking utensils. It's sort of like, tries to figure out all the different information that might be useful to help. And so we got a course like design tips, light colors and using vertical space. But then something called a lift top coffee table, which apparently is a regular coffee table, and it's got storage underneath, and it rises and lowers to eating height. So, you know, when you want to eat, it comes up and it's a table size. And when you go, where was that coffee table? Right, yeah, exactly. And it also anticipated what's next as we followed up. And it offered first local stores with inventory that had that and other ideas for, like, small kitchen appliances. And so those are all opportunities that, you know, in the old days of search engine marketing, you wouldn't bid on, you know, tiny apartment ideas.
- [18:31]
B
Right. That would be super.
- [18:33]
A
It'd be two upperforming it.
- [18:34]
B
Right?
- [18:35]
A
Yeah. And so the user wins by solving the problem. The advertiser gets these new moments of relevance to be there. And so that, I think, has been the, as I mentioned before, the big aha moment and why Advertisers, I think, are really leaning into things like AI Max and performance Max to. I can't come up with a keyword list to address that use case that I just gave you, but I can use AI to understand what are those moments that are relevant for my business to be there.
- [19:03]
B
Theoretically, this brings more search advertisers to the table. It's. It's still very early and it's still unknown, like, how these things will perform. Do you almost have to, I don't want to say, train or reset expectations for how performative search advertising would be because of these different moments that are maybe less like, I'm buying something right now versus I'm not sure.
- [19:23]
A
Yeah, if you. If you take this notion that, like, you know, searches are going to be in different parts of, you know, the funnel, as they say. Right. Can you have the same expectations for all of those? I mean, what we're seeing so far is a great product fit for the AI features in search, and advertisers are adopting it. AI Max is probably one of the most successful launches I've ever been a part of in the ads business. And overall, we're seeing monetization right now in These new surfaces at approximately the same rate. And so advertisers are finding those moments of, of relevance. We've also launched some new tools to help advertisers figure out how to bid in these kind of multimodal experiences. And so search bidding exploration is something we announced earlier this year to help kind of experiment with like what's the right bid for some of these queries that might be a little further afield. And so it's early but I think advertisers are embracing it as opposed to like ah, that's not really my department, that's like the brand advertising department and you have things like that. So. And I think the purchase journey hasn't been linear for a long time. And so advertisers are kind of pretty in tune with understanding things like well, I want to be in these moments that matter. I want to understand, you know, does even being in these moments, do they drive any brand value for me in addition to whether they drive any.
- [20:47]
B
How do I capture that when I measure it?
- [20:48]
A
Right, yeah, yeah. So again I think it's relatively early, but we've been, we've, we're on a great foundation where it's monetizing well, users are searching more, they're finding a lot of benefit in the experience that's creating new commercial queries. And so the flywheel there seems to be moving quite well.
- [21:06]
B
Any other examples that you can provide, like experiments you've tried or different brands that have come to the table and found new kinds of queries that have been working for them.
- [21:14]
A
Generally speaking, the experiments that advertisers are doing is generally adopting AI Max, adopting performance max, putting their entire product catalog if they're, if they're a merchant, more creative than them. I used to see there's a lot more creative generative. AI Creative has been. One of the more exciting things that I've seen is the ability to scale Creative from your product feed, to be able to customize it and things like that.
- [21:43]
B
And so advertising, they're not cautious. Like I think some big, big brands are still seem a little cautious about going all the way with generative. But you're seeing these search advertisers go for it.
- [21:52]
A
There are there, I mean search advertising, the canvas is, you know, relatively straightforward in a lot of ways. And so we're finding two things. First of all is the majority of advertisers are doing some form of asset based creative today. Right. They're providing, you know, text creative landing pages, visual assets where we think they might be additive. The Experience. Right.
- [22:15]
B
We're not talking about commercials just yet.
- [22:17]
A
What's that?
- [22:18]
B
We're not talking about TV commercials just yet and having somebody replace everything.
- [22:22]
A
Right, right. And so in the, in the video and the display advertising space in particular, there's a different level of appetite. And so we have Asset Studio and Google Ads that allows you to create all these different visual and sight, sound and motion creatives. But you know, advertisers have the ability to be like, well, you know, I don't want this color scheme or you know, you can, you can put some guardrails on it, brand suitability and things like that and, and also approve the creatives before they're deployed widely. On search side, there's, there's again a much more healthy appetite to experiment because of the canvas.
- [22:58]
B
You've measured product catalogs a couple of times. We've seen all, you know, like there's a lot of excitement around. Oh, Walmart still has shopping ads and chat ggp. What are you seeing right now with shopping? Does it all go to the bots? Does this totally change retail media? Does this just become another place to shop? Like what, what are you seeing early?
- [23:14]
A
Yeah, I think, I think the Walmart ChatGPT. I'll leave it to them to talk about that. I don't think it was an advertising announcement, but I think it was.
- [23:23]
B
No, it's not yet.
- [23:26]
A
An opportunity to kind of, you know, facilitate a transaction when, for example, you know, like we talked about before, where the journey puts you in a place where a product from Walmart might be the right answer to what you're looking for. It's a great example, I think, of how merchants are realizing that this changing use of behavior around gravitating to AI powered experience, let's find the opportunity to add value there. I think to your point, it's early agentic, we think is additive to these tools to help businesses drive better results and with less effort while also helping people get stuff done. I don't think the core tenets of marketing around connecting audience to products and services are the right moment, but the tools are definitely changing. You know, a good example of this, you know, at I O we announced an agentic checkout feature to help people track prices, set how much they want to pay and then buy the items when the price is right. So I always use shoes as my example, but if I want to buy a pair of shoes, but I only want to buy them if they're at a certain price point, you know, instead of losing me as a shopper when I'm in that moment. It's like, ah, these are too expensive or I'm going to wait for the Black Friday sale or whatever agentic can just find the price drop and then retailers get the win. And so it's like, you know, buy from me when it hits this price rather than me abandoning the shopping cart.
- [24:45]
B
You don't have to sit there and stare at the machine and you know, poke around.
- [24:49]
A
Do this for me. I'll take it off my to do list and then that helps the user get what they want and the retailer critically doesn't lose that moment either. Right, right. Yeah. And so I think that I think is a probably a key insight in what a lot of the merchants are thinking is like how, you know, how do I get in these moments to reduce the friction? Like we think about this even in search. Like anything that we can do to reduce the friction of people searching we're trying to do because we find that when we do that people search more. Right. Think about, you know, people taking pictures to do searches with their phone camera 25 billion times a month. I mean those are the types of things that those weren't search queries two years ago.
- [25:28]
B
There wasn't an opportunity before.
- [25:30]
A
I think the merchant space is similar. It's like how do I reduce the friction to get from start to end of a purchase journey? And I think the real power there lies in how this comes together in the future, how agents and advertising can work together again to make it more seamless. So I think that's the macro trend.
- [25:47]
B
Last thing is I was thinking about this. What is people are talking about Google facing search competition from. From OpenAI specifically. I know there is perplexity. Is that your competitive set right now or is it also these other search vehicles like TikTok, like Amazon, like Pinterest, what is that all the kind of this same, same area that you've got to focus on.
- [26:08]
A
Yeah, look, I'm in the ads business. So anywhere people are spending time and marketers are spending money, that's competition or addressable market for us. And so to your point, you know, AI is expanding what's possible and it's introducing new ways for people to discover and new user experiences around that. And so it feels incredibly competitive right now. Fortunately, I also think the pie is getting bigger. It's not a zero sum game, but even just think about the advertising space. In the past few years alone we've seen countless new entrants even outside of AI. Right. Like I'm competing with grocery stores. I'm competing with ride sharing services, big box streaming services.
- [26:46]
B
Everybody's an ad network thing. It's, you know, everybody's got an ad.
- [26:48]
A
Model now and then from an AI perspective. And certainly there's a lot of chatter around that. And so there's the consumer thing where we want to think about competing for ways that people find and discover information and the advertising opportunities that go with it. But again, I think it doesn't feel like a zero sum game, but it feels incredibly competitive today. Yeah.
- [27:10]
B
All right. Well, then good luck with the next 25 years or so. Awesome stuff. Thanks for your time here. Let's talk in.
- [27:16]
A
Yeah, Thanks a lot, Mike. Take care.
- [27:18]
B
A big thanks to my guests this week, Google's Dan Taylor and my partners at Elemental TV and sabio. If you like this week's episode, please take a moment to rate and leave a review. We have lots more to bring you, so please hit that subscribe button. We'll see you next time for more on what's next in media. Thanks for listening.