 So it's been a great two days of content. I hope you all will agree. And that's largely because of our sponsors, MIQ. And I'd like to introduce you to Charlie Nier, SVP of MIQ. Thank you. So I think it's important to tell you guys what we're going to get from the session. And if you want to leave, I won't be offended. I'd rather you stay. And if you want to text more people to come in here, that'd be cool too. So a couple things we're going to talk about during this presentation is, how should your teams interact with AI? I think Kathleen just a really good job hinting at some of the ways that her company maybe could use AI, or could use HI, which we're going to talk about as well. And then also, how does AI work with programmatic? I don't just mean display media. I mean, anything that's bought, transacted with, or auction that you guys are going to have inside your media mix, social, OTT, programmatic TV, whatever. I'm going to give you guys a couple best practices around what we've experienced with AI. Just to let you guys know, we have 300 master's level and PhD level data scientists and people who study artificial intelligence. And they've chosen me to come up here and talk about this, which is a big mistake. Just kidding. But if I let them down on this, I won't hear the end of it. And I know they're watching in India. So hopefully you guys will find that part valuable. And then lastly, some practical examples, like things that we've seen and learned from in use cases. OK, so before we get into AI, I think it's important to talk about technology, because AI is a subset of technology. What technology does is automate human behavior. So I was sitting around my house working on this deck the other night, and I was thinking, what is something that is technological that has saved me tons of time? And of course, you can look around and see everything. And I thought the remote control, the amount of time that remote control has saved me from walking up and changing the television channel or trying to manipulate my TV was massive. And that's because technology automates human labor. That's what it does. That's what it's designed. That's part of the definition, et cetera. What AI is going to do is it's going to automate our rationalization. It's going to automate how we make decisions and how we ask questions and start to answer some things that will free us up to think about the next challenges. So any of the reading that you do on AI, it's going to talk about machine learning. It's going to talk a lot about data inputs. But it never really says what it's going to do. Like, AI does replace our rationalization. You see it when you log into Netflix and you see the suggestions. You see it when you go to Spotify. And you can notice that what they select for you to play is getting better day on day so that you don't have to make as many decisions so we can move faster. So I'm going to give you guys a story. I was instructed by my marketing team to be personal, which is a struggle for me, because I lack empathy. So I'm going to go for that. So this is not my house, or that would be creepy. But when you wake up in the morning, I'm just going to tell you guys a quick thing about my day. I'll show you why it's important at the end. Wake up in the morning. I check my phone, New York Times Crossword. Usually I get jumped on by one of my kids. I usually oversleep, so I've started getting in a hurry. That's how my typical morning looks. There's a couple data signals in there, which will come back to you. I commute. I live in Chicago. I ride the CTA. Those are ungloved hands, which don't exist in the winter. But so this commute is important because my geolocation changes, and also my mindset changes. What I'm looking at, what I'm experiencing, the people that are around me aren't my usual network and aren't my usual times of travel. This is not our office, but we do have more room to work, not much more, but more. But at work, we all use desktop. We use our phones. We're online. We're getting data signals. We're reading brand week blog posts, hopefully. We're doing all the different things that we do, and we're collecting data signals as we go. Lunch. Hopefully if you work at a brand, you don't have to have a rep lunch with someone like me, or you can be pretty bored. Usually you go by yourself. You go, you eat. Again, location changes, data signals. Maybe you order from Caviar. Maybe you order from Uber Eats, whatever it might be. Then it's time to go home. Not after lunch. That'd be a typical sales guide day. Usually there's more work. I guess we skip that slide. But then you would take a Lyft or an Uber. Maybe you go meet your friends. So you've logged into your phone. You've had an application usage. You've chosen a destination. Maybe you played music on the trip. Maybe you split the fare. All of those different things are data points. You get home. One of my favorite questions is the what do you feel like for dinner question. That's never fun. I spend a lot of my time on my Uber home googling recipes. My wife knows what we have in the refrigerator. She knows what we want to cook. We start to coordinate that. We get home. We do that again. Data signals. This is me being totally unabashed by admitting to you guys I Google parenting tips a lot. If you're a parent and you don't, you're a liar. So I'd also like to say that if these were my two children, that would be my son and the heir. And my daughter is definitely Darth Vader. She's a little bit harder to deal with. But the important thing is there's this time in the day when I'm trying to seek knowledge and gather information and try to better myself. And a lot of times that happens after dinner. Then we put the kids to bed. So after we put the kids to bed, my wife and I have a conversation, which is usually just me going online and doing things she tells me, paying bills, checking insurance policies, booking flights, ski trip, whatever it might be at that time. Then the ultimate debate, which one are we going to watch? This has been used a lot this week. But again, data signal. We log on to Apple TV. We're cord cutters. What we choose matters. How much time we spent on Hulu during Handmaid's Tale season versus how much time we spent on Netflix. Wondering why that one gentleman was willing to go to the ends of the earth for this music festival. So that's an important time in our day. So before that, then after that, the intimate moment in bed where we both get on our phones looks remarkably like this. A woman actually looks a little bit like my wife, which I didn't know until right now. So what does all that look like to us as marketers, to the humans that we are as marketers? We start to bucket those people. We have a term for that. We all know that term as personas. So why do we use personas? It makes it easy to batch users up. You guys are running on limited resources. You don't have all the people and technology you necessarily need to segment or fragment your audience as much as possible. But I'd be willing to, really willing to bet that besides male and female, almost everyone in here is sits in that persona that I just described. Like a lot of our days look remarkably similar to that. Like we all get up, we all go to work, we all eat lunch, we go home, we take Uber, we take Lyft, we watch Netflix, we watch Hulu, we go to bed. There are different proclivities there that make us unique and make us different. But building personas is a limitation of human thought and the ability to process a lot of data signals. And the problem with that is the likelihood of getting exactly what you want is small. It's kind of like a gumball machine. I think about this one a lot when we talk about audiences and impressions. It's like you go up there, you guys have money to spend. You put your money in and sometimes you're like, yes, the red one. That's my favorite flavor. Probably 95% of us prefer the red ones if you don't, probably some serial killer test for you online. And this is important because there's a big potential wastage. There's big potential wastage on not getting what you want, not the right impression on the right inventory to the right person, being in the right moment if you're a brand, serving the right GIF or having competitors GIF when one of your loyal customers comes up based on the presentation yesterday. So that's how humans would look at all of those data signals that I collected throughout my day. So there's another element to this. How would this look to technology? This is really important because technology does an awesome job at organizing data. It takes all those data signals. So it takes my waking up and hitting the New York Times crossword. It takes me Google searching parenting tips. It takes me using Lyft in my destination and I split. And it organizes. Technology does the job of organization. It organizes that data to where it needs to go. It goes to CRM systems. It goes to display partners. It goes to Facebook. All of these different elements start to go different places. But that has limitations. Organization doesn't focus on outcomes. Like, yeah, it's cool that Hotel Chain might have a ton of CRM data. That's super important. But what does it mean about how they use it? That's why we still have to use humans. It doesn't determine an outcome. Just because you have your data centralized and you know a lot about your customer doesn't mean you're giving prescriptive outcomes on how to get more of them or go from 2.1% or 2.1 nights average state at 2.6, 2.7, things like that. They don't focus on those outcomes. And then how do all of those data signals look to artificial intelligence? That's really why most of you hopefully came to this room. I'm gonna go back to that slide. It looks like individual real-time decisions. Like, if we wanted to and the government didn't intervene, we could give everybody in here essentially a digital serial number. And I like a barcode and scan everybody and say, hey, Charlie, no, you went to the doctor today. Is everything okay? That'd be creepy. We shouldn't do that. But that's what AI can allow us to do. That's how powerful it is now. And something that's important is people think about that being the future and being scary in the future. The reality is is AI of Google, Facebook, WhatsApp, Apple, they're all attracting all of us right now. So if you're fearful of AI, I'd recommend getting over it because it's happening right now. In this room, we're all being tracked by these companies. But there are problems with AI. It looks beyond one problem. So if you think about the gumball analogy, like getting the gumball you want, the gumball being organized. AI, and a lot of times we see this in our system, it comes back and it tells us the wrong problem. It says this person shouldn't be eating gumballs because they have cavities. They're searching for dentists. They have small children, they shouldn't be eating. All these things that we never even thought a lot of our technology would be able to provide, it starts to ask questions back. Context can be lost. So there was a Microsoft chatbot. Have you guys seen this on Twitter? That Microsoft released a chatbot on Twitter. And it was all AI based and within, unfortunately within like 20 hours, it had been turned into a blaring racist. It was copying like Trump supporters tweets and tweeting them out there. Because it thought that that's what it was designed to do. And it started to take this on. So the context of the social environment can be lost by AI. So for takeaways, this is the first one. Really selectively placed. I was really nervous about this picture because the first one I saw it, I thought it was the other finger and it would be aggressive. Our first main takeaway from this presentation is, and actually we hit on it in the last presentation, is AI isn't the be all end all. Like you have to combine this with human intelligence. That's what we call H.I. And with technology, that's the routing and sorting and processing of information to make better business decisions. So that's the first takeaway. The second one is really about transparency. The companies that we all work with, show of hand who buys with Google. Probably everybody. Okay, yeah, Facebook, same thing, right? All right, so pretty much everybody does that. You don't really demand to know from Facebook and Google how they're making artificial intelligence decisions about your media dollars. I'd argue you should. That you should know you're in control. You should know that the things that you're looking at have been customized to your business. That they're optimizing towards your KPIs, and I'm gonna talk about that in a minute, and not towards the control and the betterment of their organization. Be ethical. This is a really interesting one with artificial intelligence and something that you guys should start to think about as your data science teams come to you and say, hey, we wanna try this. We have this new possibility to track users this way. There's a lot of legislation being passed right now that is saying any artificial intelligence algorithms, any type of action that might be taken by, let's say, a self-driving car is the responsibility of the company. So the AI technologies that you guys produce and use will be your responsibility, and it's something that we've actually seen some of our UK partners being a UK-based company have really struggled with because they're way tighter on privacy and what can be intruded upon upon a user. And this is the most important one that we found, and I talked about it in the first point. It's really about customizing artificial intelligence, human intelligence, and technology for what you guys do. There are a bunch of really good examples of things that we've done, but really it's about demanding your KPIs. All of us is challenger brands. If you're Casper and you're Purple and you're using the exact same inventory sources, you're buying on Google and Facebook, you're buying Instagram, you're buying AdWords, et cetera, you're using the exact same technology stack. So the only thing that can separate you is how much, and you're gonna have same users, the only thing that can separate you is how much you spend. If you're a CPG company and you're, let's say, new to the market, the only thing that differentiates you from, say, a P&G or a Unilever is your ability to customize your technology stack because if you try and use the same technology that they are, you're gonna get priced out and beat out of the market. And I think that's super important that you keep in mind your business KPIs need to be the center of your artificial intelligence. The things that you guys care about, that average stay one from earlier, that really kind of resonated with me, we're gonna talk, because I'm gonna give you an example of how this looks in a minute. I feel like that's a business KPI, but my guess is when you go to buy and place media with a KPI like that, it's tough to go to all your media partners and say I need to increase my average night stay from 2.1 to 2.6. They probably go, that's cool. Here's a whatever dollar CPM it is and that doesn't really help you. So whenever you're setting up AI, make sure that you demand your business's metrics, KPIs with your partners. AI is smart enough to incorporate those, whatever the KPIs your business are, if it's acquisition, new customers retention, AI is smart enough to ingest those if you set it up in a custom way. So I think to the example that I just heard and I'm gonna try and do a live example which is always risky, on how would we would combine AI, technology and AI? I'm gonna start with technology. Technology would be last week, Chicago's weather was really bad. I went outside, my face froze. I could literally feel my face freezing. It was miserable. I went to a hotel. So in that instance, it was a Kimpton. If we had technology set up correctly, the representative would know that like, okay, this person is coming from a cold weather climate. How do we make them feel warm and welcome? Is this a cold weather environment? Is it a beach environment, et cetera? The AI would be the person at the front desk being able to decipher what to do with that information. It's like, okay, this person, we're in Miami. This person's come from Chicago. Let's make sure they've got sunscreen in their room and then note this as, hey, feel free to hit the beach. Here, instructions to the pool, things like that. Maybe if they were going to Minnesota, it'd be like, hey, you've made it. Here's a warm tea or whatever it is on us. So that's the AI. But the AI is really the important thing because this is gonna tell you what other questions to ask next year. Like, hey, this year we saw a big increase in retention because we made these changes to how we interact with humans. What other questions could we be asking? Could we be featuring our bathroom pictures on our pages more? Like, that's the type of thing that AI can help you do. And it can help you generate creative on the back of that. And it can help you segment markets and do all the things you would do with infinite resource. So that's an important one. I would really want people to ask me hard questions because I did a lot of training and preparation for this. So if there are computer scientists and data scientists in this room, don't raise your hands, but anyone else who has a really hard AI question, I would love that. Anybody? All right, cool. That was useful. Well, we got one. Wow, thank you. Do you recommend any resources for people to stay up-to-date in any kind of legislation or best practices in terms of personal data for customers? Yeah, great question. And that's a really good question to ask us in particular. So we are a British-based company. The UK government was one of our biggest clients and is one of our biggest clients. So we had to be one of the first companies that was totally GDPR compliant. My recommendation is, first of all, hire experts, like get someone in-house that reads the Compliancy Law and understands everything. If you're a global brand, you're liable if your customers go to Europe or if European customers come here. So understand the legislation. There's something coming called CCPA, the California Consumer Privacy Act. I don't know how many of you guys have seen that. That represents around 12% to 17% of the US population depending on how you count certain demographics. That is gonna be more strict than GDPR. So knowing that's coming and knowing the things that you're gonna have to do are important. And a lot of this and a lot of the recommendations I would make around this are you have to have all the information and the best policy is to trust no one. Like, you don't put any information that could be questionable into any system. Like email addresses into Facebook without hashing them. Sharing segments across partners that could potentially contain PII. These are all things that are gonna be big liabilities next year. Arizona I think is the next state passing a privacy law and these laws will protect users that travel from California to other states or from Arizona to other states. So you have to understand the law and set up your technology stack so that your media teams and your advertising teams don't necessarily have to worry about it. I got a bunch of material I'll send it over to you later. Any other questions? 52 seconds. All right, cool. Thanks, Lisa. Thank you, Charlie.