 Hi, good evening. I really don't know what I'm doing in a group full of marketeers, because I think the CEOs probably are the worst enemies of marketeers, so tells my CMO to me. And then on top of it, you know, a CEO with zero background in marketing who looks at everything in marketing and says, what's up, take care, but please tell me what's the value you're going to get from this. So what I'm trying to do here is I'm not being a tech expert, I'm not being a marketing expert, I'm actually trying to give a CEO perspective to a roomful of people who will probably listen to me without frustration because I'm not impacting your budgets, I'm not approving your proposals or otherwise. I wanted to bring to focus some of the incidents that are some of the experiences that we're having in terms of using AI in marketing and I think people have been talking about mobile marketing, expertise, data, crunching, analytics, third-party data, first-party data, I think all of you are the experts, so I'm not even going to touch that area because you know best what all that is about, but what I do really want to talk about is look at it from the perspective of the CEO who's actually signing off the budgets. I just wanted you to step back and see whether there's any way you will be in a position to address that particular objective. Look at it from personal opinions. At a personal level, I think AI is fascinating. As somebody who started working in, and yes, I'm not all that ancient, but I did start working in a pre-digital era, we actually had ledgers. Any of you know what ledgers are? I've seen them in some Google search somewhere, but yes, we used to actually work with physical ledgers in banks with ways to have notes when you had to do a consolidated balance sheet. It meant taking such a huge paper with columns and plotting numbers and totaling them. So that was the era I started from. So looking at the way digitalization has moved, computerization has moved, technology has moved, has been fascinating, because I didn't take anything for granted. Unlike all of you and unlike some of the people who are born into digital, and what was it that he called them? The previous presenter? What did he call that? Trujian. Trujian. Unlike the Trujian people who probably take a lot for granted, I take nothing for granted because I actually know what it was like to have nothing. What it was like to have to coordinate with your boyfriend before you leave your house, before he leaves his house, because you only had landlines. What it was like to miss a bus and then figure out whether he's going to be waiting for me or am I going to be reaching there without him standing at the bus stop for me? So I take nothing for granted and I'm extremely appreciative and I'm as fascinated with AI like a child is with a toy. Having said which, you have to put on your pragmatic hat, right? And then you put on and say, okay, all this is fine, but how are you going to be getting any results? What are we seeing? You know, what are we seeing in terms of the way AI models, so-called AI models are being implemented? And I want to really differentiate here between what we're calling as data modeling, you know, and which has got nothing to do with AI. It's got everything to do with an ability to define the tree to the last mile possible. And as detailed as you can make a tree, it's that detail is the app that gets generated. It is not AI. You can put in machine learning there on top of it and then the machine learning is a whole different ballgame altogether. Are we really using AI models which are interactive in a real time is a big question mark that I've got. Are we actually spinning off, you know, just data models that are built on logic and tree as AI? I see more and more of data models being spun off as so-called AI and not real AI. Real AI, I think, you don't have. If you really look at it, and again, I'm telling you, this is not a tech view. This is not a marketing view. This is somebody who's looking at the ROI, somebody who's looking at what is the bang for the buck that you're getting and that perspective. And I'm giving you that perspective and I say, okay, I've got this whole bunch of rules and then you define it and then you have some more. We're talking today of on one hand hyper personalization where we say, you know, the 50 people in the room, each of you have got a different preference. You may be of the same age group, but you'll have different things. And then we go and put an entire Trojan community and say all Trojan is going to behave like this. You know, no digs at you, but I'm just trying to figure out, you know, how is this entire dichotomy being managed? And that's where I genuinely believe AI is a solution where you can actually interact with the individual, see and categorize them into their own unique segment. There is no group that you're categorizing them into. I mean, data just does not mean that if you had an earlier, you know, classification in marketing, which is probably broad at three levels on, then it's became nine levels, then become 18 levels. Now it's become 300 levels because you can model that much. I don't think so. The moment you're categorizing anybody into a model, honestly, that's not AI. AI is about the ability to personalize at that individual level that you're talking about. What's the big other concern that I have when you're talking about technology and marketing? Look at the two groups, right? Look at all your data scientists. How many data scientists in this room? No. How many marketing people in this room? The whole bunch of you, because you're social, because marketers like talking and meeting people and interacting. They find human behavior fascinating. They like exploring the emotions that go into it. I mean, they can create a huge difference between Pepsi and Coke, but at financial services, they struggle to create the difference between a participating and a non-participating product. And I bet none of you here know it. I just want a bit. Oh, you know it? Is there anybody here who understood what I said when I said power products and non-power products? No. So I just want the bet. So I'm just telling you these are two broad segments of life insurance products that are available in this country from time immemorial, OK? And they're completely distinct from each other, totally different. But we won't be able to create the distinction. But we can create the distinction between Pepsi and Coke. Because you're not selling on facts. You're selling on emotion. Am I right? Now turn it around to data scientists. What are data scientists doing? They're only crunching data. They're only crunching data and crunching data and then going around giving you prediction. The more data they have, the more prediction they can give you, the more prediction they give you, the more precision they give you. And the more precise they are, it means it enables granularity. It enables that all of you as marketeers can decide whether customer A is charged differently from customer B, whether a green color suit is charged differently from a dark blue suit, and so on and so forth. I am told, and you all must be already aware of this, that the entire Uber pricing is on AI, right? It's all on the base of data analytics, not AI. Sorry, I take back my word. It's all on the base of data. Now if data scientists are doing this level of change, are marketeers geared up to really move with that level of detail? Are you in a position to change every communication point, every single way you talk to the customer on the basis of this granularity of data that you're getting? Are you able to then tone your messages on the base of the fact that a 25-year-old in Calcutta at 10 a.m. will respond like this versus a 25-year-old in Bombay at 11 a.m. will respond like this? And can you change your communication? Because that's what data scientists are giving you. Data scientists are giving you that last mile of granularity. But I don't see marketing being able to really gear up like that. I don't see product being able to gear up like that or pricing gear up like that. And then I wonder, are we trying to put the cart before the horse, because do we even have the capability to really take so much of data analytics, so much of insights, and do something different with it? Or are we taking all of that and then dumping them into two or three broad segments that we understand? So that's a big question for me when we're talking about, because none of these are cheap. Data science modeling is not cheap. And if you get it wrong, the cost of misalignment is very high. So you need to make sure that you got it right before you deploy it. And then after you deploy it, the ability to communicate back to the data scientists, every single movement of the customer so that they can then make corrections in their modeling and redeploy it. And then you again test it out and then you send it back again. And both of you are so different. Data scientists are so different from market years. Do you guys even have a common language? You know, we're talking about, I come from the financial services. I work in life insurance. And I've been in insurance from 87. So insurance is something I breathe. And I feel insurance is so alive. I haven't found a marketer who is able to create differentiation in insurance between the 24 companies. You can pick any ad of any company, take out the name, and it could be anything. You could replace it. That's how, what do I say? Non-creative, our products are. They're so challenging for marketers to create anything different, right? So I said, all right, you know what, forget it. We're not going to use marketing. We're going to try and do AI modeling now because individuals can appreciate it. There's somebody, there's a young mother who's going and buying stuff from first cry. She will definitely be open to thinking about what if, are you guaranteeing your certainties? You know, you are going to live and your child is going to grow. Are you planning for your child? She'll be open to it, so I'm going to go talk to her. But do you even know that every young mother wants to talk to you? No. There are probably a lot of young mothers who are like, please, can I just manage my baby and me and my postpartum depression, please? I don't want anything else right now to be discussed about. How are you going to get those nuances? And that can come through with a lot of data tracking and data analysis. For sure. And the misalignment, the communication between both the technology team and the marketing team, I think that's something that I really want to draw attention on. If all of you are planning to work on a lot of data, please make sure that you find a fantastic two-way communication with your technology team and your data scientists because if you don't have that, you can come back and visit me. But I quite guarantee you that none of your models are going to be successful because the amount of feedback that's required to tune those modeling is not a joke. The last three minutes, I'm going to talk about one more since I heard the earlier panel, and I was hoping really to hear the other panel also, which was supposed to be a complete tech team. And I thought I'll hear this marketing panel. My big concern in the way people are brand marketing today is I don't think we are measuring the dissonance. All of you today probably have measurements of the effectiveness of your models, right? Right? You know the measure of measurement through clicks. You probably do it as a social media to hear customer wise. You're probably doing the mobile on track. You're doing the measurement using user engagement. How many people are actually, you know, what you showed beautifully in your slide? Are just blocking out videos. They're not looking at ads. They're just moving on. And how many of people are just allowing it to go on? How many of you have been chased, literally, after you've gone to a website? All of us, right? We just get chased. And I'm like, oh my god, this guy is wasting his time and his money. Why is this brand wasting their money? I just went and bought five pairs of dresses. I am not buying right now. Why are you sending me new dresses every hour of the day after I have bought five dresses? And when I see this kind of marketing efforts going on and I experience them as a user and then I come back and put on my CEO hat and say, I don't want to do this because I don't think I'm unique. I think I'm the norm. I think if I don't like it, I think a lot of people won't like it. And if this is what mobile marketing is all about and this is how data is being used, this is how first-person data, that website had first-person data. I mean, in this world of data privacy, do we even know how valuable first-person data is? And you're taking that valuable first-person data and converting someone who was your brand ambassador into someone who's never wanting to come back to your website again. And there is no measurement of this at all because there's no way anybody is tracking how people are reacting to this constant following. There's a word for it. I'm forgetting the word, but it's that whole retargeting, right? How are people reacting to the retargeting? And there's no measurement of that at all happening today. So I think with this, I will close my speech. I hope I left a lot of questions in your mind. I, as I said, I'm like a child with a new toy as far as AI is concerned, but when I put on my pragmatic hat and I don't see the benefits as yet, I think there's a lot more work that marketing teams and tech teams have to do together for AI in marketing to be a reality. Thank you very much.