How AI Is Leveling the Playing Field in Marketing

(00:29) I actually met him through another guest of ours who's on the call today, Mark Coleman and old friend and and our paths across so many times of I even saw him show up in my building one day because I was neighbors with his daughter and uh and my daughter went to school with his grand daughter.

(00:48) So all kinds of connections actually brought Jeff here and when I saw Jeff speak at an event that was hosting and and beyond Jeff's amazing title and work I was inspired by some of the work like you know I I think sometimes I'm seeing a lot that's out there. I always know I'm just scratching the surface.

(01:15) And even in work with Google and they they know a few things about publicizing work and making sure uh brands and agencies know about it. I learned a bunch of new things and so Jeff, so excited to have you share this with uh more of our community and and so we can all learn a bunch of new things from you today. Welcome. Well, thank you so much, David.

(01:37) Thank you for um inviting me to join today and you know it is funny you talk about all these connections and so on even with AI and so on there's you know the world is about human connection and all these different uh different people and I think one of the cool things about AI and just the waves of technology we've seen over the decades or centuries is that hopefully it frees us up to do more things that we love and to make more connections and so on.

(02:03) And so I think that's really the the most important thing. So but anyway, thank you for having me here today. I've been at Google uh for close to 18 years. Was a double click before that and Apple before that. So I feel like a dinosaur, but it's um it's it's super exciting. The world is constantly changing.

(02:24) People say 18 years at Google, what must that be like? And I sort of liken it to that movie 50 first dates because it's I feel like I've had 18 first years at Google because every year it seems everything changes. And so anyway, that's sort of the scoop. The title of what I'm going to talk about today is the new level playing field because that's really what I think it is, okay? is that AI is uh providing us and I'm going to be at Penn State my alma mater tomorrow.

(02:55) Um they're not having the greatest football year but in truth um it is the new level playing field. This is a shot of Penn State's field because in reality um as anxious as a lot of folks are whether you're 18 or 80 I know there's a lot of people that are anxious out there. Is AI going to take my job or is a going to take my grandson's job or whatever.

(03:20) The real story here is that you know AI present any anytime there's disruption like this whether it was the internet or mobile or the printing press 400 years ago there's disruption and it creates tremendous opportunities um and granted certain things do get disrupted and go away but there are tremendous opportunities which we're going to talk about um because and I think those folks who embrace it and are scrappy and say you know what everybody's going to have access or does have access to the same tools. Now, so what is going to differentiate me from

(03:52) somebody else that has it? Or if let's say you're in the lumber business, you're 84 lumber versus lumber liquidators versus Lowe's versus Home Depot. What's going to be the difference? How do you differentiate yourselves? And it's that because you're all using the same tools that's going to be the true differentiator.

(04:10) I want people to feel uh open here to to asking questions and so on. this should be a conversation, not me just speaking all the time. So anyway, that's sort of the way I I think about it and why I'm saying the new level playing field. Um, so it is the new level playing field. Everyone has the same access and your skills and this is, you know, everybody's skills in collaborating with it will set us apart and will set each individual apart and each company apart.

(04:41) Um, and so before we get into that a little further, I'm going to talk about what I do, my passion, uh, in addition to work is I'm a sculptor and there's definitely a strong AI component to sculpture. Um, I don't know if anybody is familiar with this sculpture.

(04:59) This is a piece I did that was in Riverside Park uh, a few years back at 145th and the Hudson River. It started out, the right-hand image is how it started out. It was 8 in tall. It was this is what we call a Mckette or a model and I made that out of flashing that you buy at Home Depot for roofing and I this is the proposal I made to the uh Riverside Park Conservancy and then they accepted that.

(05:25) Um they loved it. Uh I was very gratified that I was given the invitation to produce it and then it went to 13 ft. Now in the process it's essentially it's a gigantic um sun dial and you'll see here that it it's aligned the shadows that's developed with looking at the shadows that are cast at the four different solar events throughout the year the two equinoxes and two solstesses and what we did was I with a collaborator I had a collaborator who did the these benches if you will the I created the main sculpture and

(06:01) then he created in green benches that corresponded to the shadows that are cast at the four different times of the year. We made this in 2016 and the process took about 6 weeks. Now with AI that process would be a matter of just a few days. So uh this just shows how AI could take a process and a lot of grunt work that really uh doesn't benefit anybody all these calculations and can condense that so that I could spend more time and actually uh fine-tuning the sculpture thinking about other things that are going to make the sculpture uh more impactful. This is an

(06:43) image of what the final sculpture looked like on the West Side Highway on the that park. I don't know if anybody passed it over the years, but there's a bike path there. It's now down in Key West. Um, so this illustrates the fact that even in my passion job, I use AI.

(07:06) And this this is an example of a new uh com rendering I did using AI for a commission down in Key West as well. I'm going to play it real briefly so you can see the type of I showed you that model that was 8 in tall. Now this with AI is how I'm going out and showing CL uh different park different sculpture parks what's possible.

(07:32) So you can see the level of detail was pretty cool much. So that's really where we're going in terms of technology. Welcome to the future of AI our technology. Let's go to the next. Welcome to the future of AI in digital marketing at so this is something I created for my presentation tomorrow and that that was done using a tool called Google flow.

(08:03) I'm not sure anybody's familiar with that but Google flow essentially I took a photo of a presentation I did and I added this prompt. Make the man swirling a magic wand while animated stars come from while he's singing. Welcome to the future of AI and digital marketing at Penn State. You probably everybody in this call and I know you're very familiar with AI.

(08:22) Uh but this these are the types of tools that I think are what are the future and just it's the sort of scratching the surface as to what's possible. Now imagine you get somebody who's doing uh playing with AI. You get somebody just out out of school who's doing some really cool things.

(08:44) they could develop a short short uh content movie or a full fulllength feature movie basically in their dorm room. You don't need to have funders and big production companies behind you anymore. You've got the ability to do this yourself or at least pitch an idea in the ad agency business. This is changing everything where, you know, obviously creators are a big thing.

(09:07) Social media, it's putting power in the hands of of regular consumers, which is really turning marketing on its head. And and Jeeoff, it looks like it's stuck on the eagle uh visual. Oh, okay. Good to know that. That's sort of I'm surprised that it Okay, that's And we're still You're still not seeing Beyond Eagle. We're still seeing the eagle.

(09:30) uh 360 view here. Okay, let me let me go back and stop sharing and we'll we'll do this again here. This just goes to show how even the best laid plans. Let me screen share again here. Yeah. So I I I Okay, so back. So basically this is Sundar's quote. Um, and it's basically talking about how AI from a Google perspective, um, in terms of search is revolutionizing, you know, what we're doing specifically at Google.

(10:07) Uh, essentially it's allowing, um, and I don't know if has anybody familiar with AI overviews and AI mode in search. So that has significantly uh if you saw Google's most recent results that has increased um the the usage there is tremendous and I don't want this to be a commercial about Google but I want you to understand why we are so involved in it and why we think it's you know people say is chat GPT going to kill Google and you know you never know 30 years ago Bill Gates uh May of 1995 Bill Gates sent a memo It's a senior level manager saying that it's entitled the internet tsunami and

(10:48) he basically was saying we need to get on board with this internet thing or else we're going to get behind the curve and even in 1995 when Microsoft arguably had the largest uh share market uh had the best brains had so much resources and the department of justice was after them if you may recall uh they still three years later after marshalling all their resources were blindsided by these two students from Stanford that eventually turned into Google. Um so I think even today no matter what Google's doing, no matter what chat GPT and OpenAI are doing, we

(11:28) could all be broadsided by some upstarts. And you know what that's evolution, but this is you know queries are getting longer and more complex. We're seeing that. Absolutely. Um and so and I everybody on this call probably the way you it used to be you'd put two or three words in now it's easily 10 or 15 more words.

(11:56) Um and that's what's sort of driving all these different AI functions back in April of 2024. And anybody on the call familiar with tokens probably to some degree but maybe not on the most technical level. So tokens and you know I'm not a I'm not developing code here but tokens are essentially the unit of measure by which AI compute is is is build if you will Google cloud builds its clients based on number of tokens they're using and tokens are basically it's a unit that a video a 30- secondond video might be 200 tokens that might require that much compute um a quick

(12:37) query that you you know say write this uh write this resume re redo my resume might be five tokens well in 2024 in April of 2024 there were 9.7 trillion tokens used across all Google surfaces and by say Google surfaces I mean Gmail uh YouTube all these various things a year later in April of this past year there were 48 80 trillion tokens used that month.

(13:11) That was a 50-fold increase. In October of 2025, there's 1.3 quadrillion, that's not a number I'm very familiar with, tokens were used. We can see the exponential growth here. And this is truly probably the best measure of the adoption of AI. Uh and this is why data centers are you know being built at such an extraordinary rate and why uh all these you know private players like Google and Microsoft and Amazon are getting into the utility business the power generation business because data centers require so much power. I don't didn't plan to go deep in that, but this this is what's happening.

(13:54) And the more that consumers get a taste of what's possible with AI, um, and I don't think we've even scratched the surface, the more that's growing. Now, you might say what's going to, you know, how are we going to sustain this in terms of power consumption? And I think equally uh in terms of exponential growth is the the technology to do this in a more streamlined power consumption way.

(14:26) So I you know that is the hope and that's where I think things are moving is that we're going to consistently see this be far more power efficient. But it's a major issue and anywhere I go and speak around the world, the hands get raised speaking about a data center that's being built, you know, within the vicinity of that venue and what how it's impacting culture and so on and community.

(14:59) I want to stop here because I feel like I've been talking a lot and I'm going to get continue on, but any questions or thoughts here? Anyone want to chime in? Well, I can continue, but I Yeah, we Well, yeah. And uh and we've got some great examples coming up. So, yeah, why don't you keep going? Great. So, I'm going to talk, you know, you mentioned AI overviews, and we had a you know, we've rolled that out.

(15:18) More than 100 countries uh are doing using AI overviews. This slide is now obsolete. As of last night, I saw a number that we're now at two billion global users every month. So, extraordinary. Um, and this is what you know it's it's going from a short query men's running shoes to how to pick the best men's running shoes with padding, cushioning, and shock absorption for a beginner runner on asphalt.

(15:43) This is the type of thing that I know I'm searching for and a lot of consumers are and this is why we have to sort of be on top of that. Um, we at Google do um and we everybody everybody's got to. And then it's especially things like when you know when you when you don't know what you want.

(16:03) That's just a specific thing as to I know what the running shoes I want are. But let's say I'm on the subway or I'm somewhere else and I see something and I want to I I don't know what it is. I don't know the name of it. All I know is I want to get more information. And this is where Google Lens and other technologies, you know, iPhone's got a camera. Everybody's got this technology.

(16:22) This is where Okay, baby. You can have a folks. Can you wait as you want? Can you wait? Can you wait as you want? Can you breathe? Can you do like this? So, essentially, this is called circle the search and it's it's the abil it's the ability to, you know, you're on the subway. This years ago, I I got a watch.

(16:51) Um, I saw somebody in the subway wearing a very cool digital watch. And at the time we didn't have this technology and I think sometimes it can be a little bit um, creepy if you do this, you know, just with somebody and you there you don't get their permission. But I had to ask this guy, where do you get that watch? And he told me the brand and he got it on Amazon.

(17:09) So I went and got the same thing. But if you see something on the street, you see something on a store, whatever, this is this is what folks are using. And 20 billion users now are using this same technology. And it's opening up a whole new level of search that we just didn't think of before.

(17:34) 25 years ago when Google started, there was that little search bar, that search window, and you typed in two or three words. Now, consumer behavior is radically changing. And once they get an appetite for it, it's changing even more. So when I talk to a lot of companies, they and that's what I talk to I talk a lot of to our largest advertisers. Um they are saying well what should my AI strategy be? And really that's not the question.

(18:00) I think we really need to get to really what is your strategy need to be that AI can enable? So, let's get back to your core focus. Whether it's if you're selling running shoes, if you're selling conferences, um or if you are, um I'm on the board of the Art Students League. If you are, uh trying to get new students to come in and study print making or sculpture, what should your should your strategy be? And how can AI enable that strategy? Because it really needs to be rooted in what you're trying to do. And then let's think about AI second second uh second

(18:40) focus. So I want to show you I want to take uh show you where AI is in terms of graphics and video. And I'm sure we're all seeing it. But I wanted to this is something cool that our ad agency did based on a spot they created for us at the Super Bowl back in February. They basically created the spot in the standard normal way with normal photo shoots and so on. And then as a fun experiment, they did it in AI as well.

(19:09) They that AI uh version did never never ran. You're only going to see it here, but they did it to sort of illustrate what's possible now with AI. I'm going to show you both videos at the same time, and then I'm going to ask you if you can tell me which one was generated with AI. So, here we go.

(19:32) Let me hopefully this will come through on the uh in Google Meets. It comes through. Let's see if it comes through here. My grandfather taught me everything I know. Our little family businesses worked their tails off. We never miss a detail and those things matter. The legacy and the tradition and the craft. It's what I live by. It's nice working with your family.

(19:50) We're a twoerson team. We're also busy moms. The financial side can be very daunting. If our schedule gets messed up, everything falls apart. We don't have a lot of time. We're not professional. with H3. Using Gemini has been a gamecher. Gemini in Google Docs helps me write website descriptions.

(20:15) AI allows us to make quick business decisions. It translates everything live. It was magic the first time I saw this. Gemini and Gmail really lets our staff save time drafting emails. It makes my response more concise and so much better. Marking up a concept or we can do that in a matter of minutes.

(20:34) Yeah, we have time to keep doing what we love and it opened up the world to me. This is a beautiful thing we're doing and we're proud of what we do. Okay. So, let me ask you, which do you think was generated with AI? Top. Any other votes? Some for the bottom here. Yeah. Okay. It was the top that was generated with AI.

(21:03) Uh I'm curious what made you say the folks who voted for the top, what makes made you say that the lighting was it was smoother? Yeah, it was more perfect looking. Yeah, it was very perfect looking. Yeah. Right. Jeff, was was the was the bottom one provided as a reference to create the top one in in certain cases? Yeah. I say certain parts of it. Yeah. So it wasn't just a prompt.

(21:27) there was an actual reference video image something used in certain cases. Yeah. Yeah. It wasn't you know that was just part it was a very infall process but yes I mean honestly if you didn't know that if you gave a script to a production team and you gave a prompt to Gemini and neither had ever seen each other then I think you'd have a very different outcome.

(21:50) Yeah. Yep. Yep. Exactly. Well, so that's the the ways some of the key ways to tell was um the one thing one of the things that at least Gemini doesn't do a great job of and I think you know a lot of the LLM same is that handling text at this point and you'll notice there was one image where somebody was wearing a t-shirt and they the verbiage on the t-shirt just didn't make sense and the other one was uh there was a child in the bottom video and Gemini will not render images of children. So at this point, so there was that's sort of two

(22:27) two cheats there. Um so let me get back and see if we can move on to the next slide here. This is this we go. So so essentially, you know, we see AI is truly unleashing the full power of search. As I mentioned, four 20 billion visual searches using Google Lens a month. Uh and notebook LM.

(22:53) Has anybody used Notebook LM? Someone's even DMing me about it in the chat today. It's like, yeah, I've got a lot of Nove fans in this room. Great. Yeah, somebody said every day. Every day. Awesome. Tell me, what do you use it for? So, um, I'm a consultant. We're part of a 140 person consulting firm and we have different clients and each client, um, it's just a given.

(23:16) If we're going to go in and do an assessment, there's lots of stakeholder interviews. All the notes for that are just the baseline for it. if we're going to go in and do branding like having that messaging architecture and all that tonality and like it's just a the perfect thing. Um and then as we go along just on the project so like it you know as you know it it's like you can do all the docs sometimes you can do a few of the sources so like you sort of want more nuance over time but it's a life ch game changer for internal project cadence. Absolutely. Absolutely. This is so know I don't need to say any more about it. This is this is one of

(23:53) Can I Can I ask you something? I'm going to put you on the spot and it's okay if this is not known or you're not at liberty to say, but in case it has been out there somewhere, at some point is Google going to combine notebook LM and Gemini so I can have like the full AI access and all the folders and organization.

(24:12) That's what I can't wait for. And then I'm like all in with notebook. Yeah, I I can't, you know, first off, I don't know that. Oh, you know, I can quite easily say it's not it's not that I won't comment. It's like I don't know that for sure. And they're smart not to tell us. I hear you.

(24:30) I hear you. It's so funny because some of these productivity tools that, you know, they don't seem like game changers. They don't seem like they're it's going to be a revelation, but this is the kind of thing where we're seeing AI inform and power a lot of these cool things. It was like Google Docs years ago.

(24:47) You know, you'd walk in and Google Docs sort of changed I think the the way of collaborating and I know that Microsoft's got the capabilities too, but it's that type of functionality which has sort of changed the way we work and you know plan trips together and do these things.

(25:05) This especially I agree with you that notebook LM is one of those things that all of a sudden it it allows you to do so much more and it's an example of okay AI giving you the opportunity to do and create and have so much more options because you've got this stuff at your fingertips. But it's the UX that works so well there. So yeah, studio IDE environment. I love it.

(25:31) Hey, you know what? You got you're you're be the pitchman here. I I have a question regarding search. Does Okay. Do people's queries affect search results or ranking? If a lot of people are asking about a certain topic or a product or brand, will that affect the ranking of search? Well, if the more that is the more there's a several different factors that go into, you know, ranking of a search result.

(26:05) Um, but one of them is the more folks that are, you know, referencing it um and find it relevant. uh and and as well the other one of the other uh factors is also other sources that link to it that reference it. It's all about sort of you know credibility and relevance and how many you know and also analyzing the content of that site and saying how relevant is that content.

(26:38) That's where the you know the the the search you know the browser is sort of crawling all the content. So it's several different things. It's absolute true relevance of content. It's then how many people are going to it and then how many other sources are referencing it. And it's sort of that's that sort of kaleidoscope of what goes into the ranking of a search result. But I got to tell you, I'm not a search specialist per se.

(27:10) I come from the other side the the world, which is more the uh the paid uh display ad perspective. But that's essentially the the rudimentary uh facts about search. Jeff, I got a question. Sorry. Um, you know how um are you familiar with Hey Jen with the you can create avatars and stuff like that. Is ever looking to create something you know how they have the voice activated uh avatars? Are they looking to have actual avatars at some point? You know that's a good question. I don't know.

(27:50) I don't know all the things on the road map for notebook LM but it's a good question and one of the things about all the products at Google one of our key innovation concepts is launch and iterate. I don't know if you're familiar with the the term here, but it's essentially and it's it's pretty intuitive from what it sounds like.

(28:15) It's, you know, like when Gmail was introduced, um it was Google announced at the beginning like this is a work in progress. Um and we're going to be having like weekly updates and so we want to hear from users about what you what you want and what you need and how we can improve this and where the bugs are. And that's exactly what we want to hear as well for things like Notebook LM. It's like, you know, we're putting it out there.

(28:38) We're sort of moving at the speed of light. Um, we feel like we are and I think we are. Um, to get this stuff out there. Um, and then but it's all, you know, like we're putting the wings on the airplane as it's going out there.

(28:57) So we want to hear from users as to how can we make notebook LM how can we make circle to search? How can we make all these things more relevant, more impactful? So I love that idea. Going to be my next question. Um how how can we be part of I guess the beta of all these things that are you know is there like a beta group that you guys um reach out to? There's not to the best of my knowledge there's not a beta group for notebook LM.

(29:23) Uh there was for instance there was Gmail was in beta for some time. So I would do the same thing that I would suggest you do is go I'd go online and do a quick search. Is there you know how do I provide feedback for notebook LM? How do provide feedback for all these different products? There's no beta group per se that I'm aware of. Okay.

(29:42) So when you go guys go la la la la la la la la la la la la la la la la la la la la la la la la la la la la la la la la la la la la la la la launch a um a specific product or have an idea there's no like focus group that you know you guys Well there may be there there definitely may be I'm I'm sure there is you know UX a lot of UX research and that type of thing absolutely um as to the process each one I think is very different depending on the product um and you know whether they've got some sort of a a beta an alpha and beta methodology or whether

(30:13) they just doing that type of sort of closed research with UX design. Everyone is super different. Okay. And you've got imagine a company as big as ours. We've got all these different groups and units and business initiatives that are out there. Um and but I would I would do a search on that.

(30:38) I would suggest you do that and say how can I provide feedback on notebook LM? and I'm sure you're going to find uh something in that regard. So, this is great. Yeah, somebody somebody put it in the chat. So, thanks. That's great. So, and this is just an example, but I'm not going to go through it because I think you guys are the most powerful testament.

(30:55) I'm presenting this tomorrow at at Penn State and it's just an example of I just put in a link to the uh the crearyy which is literally a a crearyy on uh Penn State started as a land grant college and it was an agricultural college and so there's a crearyy on campus and I it generated an 8inute uh podcast talking about the Penn State crearyy but I'm sure we've got a lot more relevant interesting stuff as well that you guys have come out with.

(31:21) I'm going to show some examples that David referenced because I think they're really cool and it comes from my background in terms of uh personalization and using uh different data sources to really drive home concepts. So I'm going to show you some things that are really net new and I think sort of revolutionary with AI stuff that just wasn't possible before.

(31:44) So the first thing, anybody familiar with Pods, the container company and storage company? Probably everyone is. Um, they came to us about a year or two ago and they said, "We want to do something cool. We want to use Gemini for something cool." We said, uh, it was part of a program. We had we had sort of put the the initiative out there.

(32:08) He said, "Anybody, you know, anybody that any ad agency and clients that want to do something really interesting with tech with with our new Gemini technology, come let us know." And they said, "We do want to do this." And they were part of our Lighthouse program where we were got a few different advertisers and agencies together.

(32:25) and they basically in the course of several months developed with Gemini um a an ad unit that uh I'm going to show you now which basically uh makes uses the pods uh the side of the truck as the world's smartest billboard if you would well and it's sort of fueled and informed this digital billboard by a bunch of different third-party signals that are using weather time uh traffic, uh, subway congestion and so on.

(32:55) So, I'm just going to play this and I think it would show you that what's possible when you completely think out of the box. Smart it would know where it was beginning here years. The pods container has been more than just a moving container driving the streets of America. It served as a giant billboard reminding people of pods.

(33:13) So, we teamed up with Google Gemini and turned a pods container into the world's smartest billboard. A billboard so smart it would know where it was, what time it was, traffic conditions, weather conditions, subway delays, and then generate hyper neighborhood specific lines in real time.

(33:40) So they did that uh over the course of the they that ran through the entire city. There were 1,600 headlines that were developed uh and you know they copywriters were involved in working with Gemini to develop those headlines which were basically uh rooted in the pod's sort of tone and manner but headlines you know based on all those different signals.

(34:08) It's the type of thing that you know whereas they might have had one copywriter working on an ad campaign in the past they had to have you know that one maybe they had like one or two actually more copywriters doing a lot more interesting varied work because this creative demanded it. So I think that you know this type of technology sort of gives us and gave pods and their agency the possibility to do a lot more and open up a lot of new possibilities.

(34:35) So I'm going to show you now results because that's the most important thing. Actually didn't I don't think I've got that here but the results here they had 66% more website visits and 33% um more signups um for you know just actual conversions. So that was pretty strong.

(34:57) So whether you're doing something just for the sheer nature of AI or whether you're doing something to drive results, obviously we're all trying to drive results and this definitely got that kind of interest. This is another example using just weather signals and in this case for Nertk. So, Nerkex a migraine um medication and what they found with this is that barometric pressure is a key indicator of uh when there's going to be migraine onset.

(35:27) And so, uh, Nerkch used, uh, weather data, specifically barometric pressure data to sort of identify and they developed using Google technology and voiceover technology different ads that were customized by region and would be triggered to run based on barometric pressure. So, I'm just going to play an example of what that sounded like for one of the key areas. And right now we're still seeing the pods example.

(35:54) So migraine weather alert attack that. Let's just stop sharing and go back to I don't know what's it's really glitchy. We we we got to clearly use meat from now on. I think that's the moral of the story. Yeah. Yeah. Um so let's go back to and if nothing else this was my uh opportunity to sell Google Meets. Um no joke.

(36:21) So, let's get back to let's go back to let's see if everybody can see that. Can you see now the Nerkch slide? Uh, yeah. Great. So, here we are on the Nerkch. So, let's I'm going to play an example of what what this sounded like. Migraine weather alert from Nerkch. Hello, Austin. There's a change in pressure heading your way. Take care. This may trigger a migraine.

(36:43) Okay, that seems pretty pretty simple, but and it it the ability to do this at scale over 200 different DMAs uh using this technology. This was something that was pretty straightforward. And then the great thing is you're able to test. And so let's say we do te several different messages um and we sort of see which one's working, which one's not.

(37:12) This is gives marketers the way to be able to do this and have a larger test and learn framework without putting a tremendous investment in. Turns out this worked really well for them. So they have continued to to do this technology. So that was just first example using third-party data signals. This next one is using firstparty data and personalization and several different and some of these are less tech than others.

(37:36) But for instance, SAS um back during CO you know and all the airlines of course were just absolutely devastated in terms of uh their business. But after co as they were coming out of it they said you know we need to to uh figure out a way to get more uh business from our existing customers and so they used uh a lot of our predictive modeling technology which by the way predictive modeling is more not so much Gen AI or it's really more old school machine learning but they they use that to sort of understand more about their customers and what they

(38:10) found was a lot of their business customers especially uh liked, you know, they had other facets to uh their travel and their interest besides traveling for work. And they so they created uh SAS created several different ads to appeal to their different customer profiles, if you will, or personas that emphasized things like fine dining or extending their trips for business.

(38:42) Um, and they ran these ads uh specifically to those types of customers and they saw a pretty significant lift, a 34% inc increase in online bookings. The ad creation in this case was not AI AI generated. uh and it's a fairly low tech example, but it is the fact of using consumer insights, using your first party data to really understand and find new patterns and really figure out how to capitalize on that.

(39:11) And then the third example or the third group is agility and personalization. And this I think is these examples are I think sort of the the wow and the sizzle beyond the the pods example of what's possible with personalization and how you can sort of do create concepts that are really cool. So this first one's for Cadbury the chocolate company that you know we have it to some extent here in North America but it's big in AMIA and India.

(39:42) This particular example is a campaign they did in India. um just coming out of COVID, a lot of the small retailers throughout India were devastated as well uh because a lot of a lot of people just did not go out and so these small retailers just their businesses were were tanked and so Cadbury came up with a way to generate uh interest and traffic for these smaller retailers.

(40:08) So they took uh probably at least in the Indian market one of the the best known actor in that market Shah Ru Khan who's a very famous movie star there and they took his likeness and his voice with his permission of course um and created technology where he basically was the pitchman for any number of different uh small retailers and I'm going to show you this example but it's all fueled with AI big brands have big money to hire famous people but not the small guys and they are the ones who are still hurting.

(40:45) Presenting Shah Ruk Khan my ad. Welcome Shah Ruk Khan. He almost certainly has more devoted fans than any other movie star in the world. Biggest movie star in the world. Forbes have called him one of the biggest movie stars in the world. We help small businesses by making Shah Ruk Khan the world's biggest movie star their brand ambassador.

(41:23) We used machine learning to recreate Shah Ruk Khan's face and voice to take the local store names in the ads. Choice of fashion say heap. Royal fashion say he MK clothes say he pass Lakshmi collection say heapro different versions of the same ad with local store names were targeted as per the pin code of the viewer showing them only the nearby stores but it is impossible to cover all the stores.

(41:54) So we gave the power to the people to create their own version of Shah Rukhan my ad. Any small business owner could promote their stores through their own social media networks like WhatsApp forwards and other social media pages. It is Shah Ruk Khan selling your store man.

(42:18) If so if you're a smalltime retailer or a merchant what an amazing this is pretty cool wouldn't you say? um the fact of being able to sort of democratize uh some of this technology that up until then you had to be a large advertiser to get a pitchman like that. So just wanted to share that. Hopefully now I'll be able to switch to the next slide. Uh but we'll see based on bottom line is not only was that sort of an impressive tech perspective, but there were 130,000 different versions of the ad created with 30 million ad views and a 35% jump ultimately the big the biggest important number metric in Cadbury's business. Um, and then lastly

(42:56) in storytelling, I think there's a big example. No marketing conversation would be complete, at least from a large client perspective, without a company like a Nike. And Nike, who arguably does some of the most coolest creative and gets a lot of views and impressions and traffic, they created something completely out of the box that was not possible using before AI.

(43:23) And they basically did it around um Serena Williams who is arguably most impressive athlete uh in the past, you know, past few decades. And they basically took footage of her um 1998 Grand Slam win and they so the 130,000 hours, you'll hear about this in a second. And they did that versus her 2017 last Grand Slam win.

(43:51) and they created a virtual match between the two versions of Serena. So, I'm gonna play this. And yeah, we're still seeing Cadbury. So, okay, let me stop sharing and we'll get to the other one. Okay. Can you see the Nike logo? Yeah. Yeah, Nike's up now. Great. So, they created this ad, which I'm going to show you.

(44:12) Serena Williams isn't just the greatest tennis player, she's the greatest athlete. We wanted to look back at how she elevated her style of play over time. Through archival footage at Advanced AI, we generated 130,000 games between two Serenas from two different eras. 1999, the year she won her first Grand Slam, versus 2017, the year she won her record-breaking 23rd.

(44:36) Let's see how Serena evolved as a champion. From the best in the game to the best ever. In tennis, there are no bigger dream matches than this one. One Grand Slam title against 23 Grand Slam titles. Here we go with game one. 1999 to serve. So, we're not going to play the whole thing.

(44:55) In fact, just, you know, I give you and I want to show you an example of what it looked like. Can anybody guess which version of Serena 1? 2017. Any other guesses? The biggest tennis geek on the call says uh the later one. Any other guesses? She altered who went one for 99 in the chat. Okay. Um it was 2017.

(45:25) So it just it just goes to show um experience uh beats um beats the agility of youth any any time. But in this particular case, this is this is how it did. I'm going to stop sharing and then go to because I know it's going to happen again. We're going to lose this. Um let me just Yeah. And and we're getting close to time, too. So Yeah.

(45:43) So, let me just I'm just going to wrap this up because we are close and perfect on timing actually. Um, is that they saw Nike saw a 1,000% jump in video views over anything else they'd ever posted. And that's Nike who's posted some pretty impressive stuff. So, this just goes to show that you know what, when we do things that are over the top, AI allows us to do things that are we've never thought about before and can create extraordinary results. We'll sort of leave it there.

(46:12) One thing I wanted to plant the seed is that there's a lot of gen Genai tools out there. I'm going to skip ahead here a second. Um but the the bottom line is there's a lot of Genai tools that users consumers have um access to. Uh but you know what they could create some pretty if you if you have a you're a marketer and you've got branded content and now consumers can do mashups of it.

(46:39) It's a whole new world which it could be a blessing or a curse. It could be a curse because if you really care about your brand guidelines, they they could be doing anything with them. It's a blessing because if you do it right and you uh think about it in the right way, you can create a whole new cadre or legion of creators and and fans, but it needs to be thought about.

(47:04) I'll leave it there. This is how I can be, you know, how I can keep in touch on LinkedIn. Any questions or thoughts here? What is the theme of your Penn State presentation? It's basically the same sort of conversation, but it's going to be more on this whole uh the new level playing field and as students or as faculty, but mostly as students, what do you need to be thinking about um in terms of being competitive in the marketplace? uh and it's sort of thinking about how to collaborate more closely with AI, how to be thinking of, you know, working with AI and

(47:44) thinking about being scrappy and using all these AI tools to differentiate yourself and when you get into a a role, how to differentiate your company from another company. One of the things we talk about here is that in that that intersection between culture and data and you know what Penn State and Nitney Lines are talking about is what's happening with James Franklin today and you know those kinds of kind of personal things that actually are top of mind and how does your data kind of address can address those fleeting cultural moments.

(48:16) Absolutely. And how do you be agile? This is where social media has become so key is because people want to hear, consumers want to hear about how Crest or Tide or whatever their favorite brand is, how they're relevant to these things. Absolutely. Couldn't agree more. Interesting stuff.

(48:37) I have a I have a geeky sort of somebody who writes about Any other questions or thoughts? I've got a question, Jeff, if I may. Sure. Um are you able to say um like how much what percentage of spend major brands are putting into AIdriven advertising and is it like mainly coming out of a testing budget or or a dedicated budget? Other words like like you know in early days digital was kind of testing and then it became dedicated 20% 25% whatever.

(49:03) Good question. I I I don't think there's one first off I don't think I could give you an answer. Uh but I also I don't think you're allowed to give you an answer but I don't think I can give you an answer because I think so many of our ad units and formats are AIdriven you know whether it be AI max or performance max or demand genen so much of it everything's got AI built into it and it and not just behind the scenes but it's key part and parcel to audience content bidding um so I mean and AI is enhancing our experience here.

(49:40) But what I but what I'm what is there a way to say like um you know for the the Cadbury thing or the Sherup Khan thing or then Serena versus Serena thing like how much are brands dedicating to to that real just sort of AI first or AI out front? AI first creative. So especially I think we're talking creative. I think the numbers still pretty pretty small.

(50:05) I would put it at probably less than 20% in terms of really AI forward creative. Yeah, it's really my area of specialty and I think that number is still pretty low, but I think it's going to grow very quickly as as advertisers realize that it's not only a timesaver, but it's also something that can distinguish themselves uh in terms of creating really cool content.

(50:37) Right now, it's everybody's saying, "Well, how can I do, you know, get rid of some copywriters?" Well, you know what the real opportunity is? How can I make my presence twice as large or, you know, have 20,000 different versions of the ad versus true it's it's the ultimate creative optimization. I understand that. Yeah.

(51:01) So, I think that that's still I think I think companies are still very much in their early stages of that. So I would put that number very low. I think there's a tremendous opportunity for us in the creative and the marketing space to come up with ideas to make this reality. I want to make Rod Waver versus Roger Federer a reality. Exactly. Any other questions or thoughts? And I'm happy to continue these dialogues down the road.

(51:25) I would love to have you know this is how David, thank you for putting this forum together because this is how great ideas come together. It's all through connection. I mean, this is like when when someone's like, "Okay, well, like, oh, you know, so much about AI." I'm like, "No, I have the pleasure of learning from so many incredible people.

(51:47) " Um, and yes, like, you know, guest speakers like yourself, but also this everyone who's showing up like I'm learning a ton from the chat today, right? I learned a ton from the Slack. So, so this is just my good fortune to have a room like this I'm in every week, let alone some of the other interactions.

(52:06) So, uh, yeah, it's like you and I get to spend a little bit more time on this than some other folks do. Yeah, that's great. This is super cool and it's changing so rapidly. Um, so looking forward to continuing the conversation. Uh, please let me know how I can help. Let's bounce ideas against one each other and take it from there. Great. look forward to sharing this.

(52:27) Uh happy to have you back, especially as you have some more ideas and examples and updates on all the exponential I mean it's crazy that I could see you like you know like one month to the next and so much has changed already. It's Yeah, it's you it's no longer 51st years, it's 51st months. So great. Yeah. Well well thanks so much for being here.

(52:49) Thanks everyone for joining and yeah amazing participation as always and see you next week.