 So we'll try our best to hold the audience. We have an amazing panel at hand here with extensive background from banking to programmatic buying, with healthcare as well. I was just discussing with Ashish what they're trying to solve for at Narana Health. So today, before I kick off the session, I have a bowl you can actually drop in your questions. I don't know if you'll have time to answer all questions, but if you could, the bowl is being passed around with a notepad and a pen in it. Preruna, could you raise your hand please? So she's gonna pass the bowl. If you have any questions regarding big data, smart data, please drop in a question there and towards the end we'll try to pick them up and answer them. And the way that we've structured this panel is we have a couple of questions, about four or five of them. The panel members are just gonna express their expertise in that. Then we have a rapid round, rapid fire questions, which hopefully is fun as well. And then we will cool off with a few questions which actually help businesses that are looking to get into the journey or have already started their journey in big data or smart data and put that through as well. I just wanna put the note there that this is a very sensitive topic for many businesses because these projects are usually confidential. So we are not getting into the detail for any business but taking the expertise view, how they may have seen other businesses or their expertise in projects that they have seen will be expressed today. Without further ado, my name is Santosh. I head the dense to CXM practice in India. With my back, I am mostly a techie. I have worked through and through in technology such as VMware and Amazon and before founding. One of my companies is Socrates. We founded that business with the notion of big data when it was not even a moniker those days. And I've overseen projects across automotive, CPG segment as well. And I'll talk to, if time permits, talk to a few case in points that I have addressed. I'll pass on the baton to Mehir to introduce himself as well. I'll just talk about his twist with big data and smart data. Santosh Till then, who gets the hamper for the rapid fire? Am I audible? Sorry, yeah. Am I audible? Yeah. Yeah, okay, perfect. Mehir is suddenly a passionate towards technology and data as a marketer which is slightly different than what probably I should have done. But my academic background was technology and I think I wanted to become a software developer at some point of time in my life before I moved into marketing. So that still remains in there and hence I touch on my tech team more often than I showed. Thio. Good evening, everyone. So I'm Uthio. I had digital marketing, my tech liabilities and propositions at Yes Bank. Been a career marketer, been in BFSI for donkeys years now. So my passion is actually consumer behavior and that's what drew me into marketing. So when I started off, obviously there was very little of digital marketing. But over a period of time, the topic that we are speaking about, data, data is behind a lot of the consumer insights that you draw and how we understand consumer behavior and that's what's drawn me more towards the digital marketing side of things. So looking forward to a fantastic conversation ahead. Thank you. Siddharth, over to you. Yeah, hi, I'm Siddharth. I had a business called MIQ. So we are into programmatic and data and my journey has been with tech industry for a long time. I was earlier at IBM and Microsoft in the tech business and in the Microsoft days, I saw all the data migrating into cloud and that's where really the big data journey started. And then I worked at Google and I was heading Criteo's business in India as well. And I realized that I think there is a very important need for these platforms to have the data layer and convert that big data into smart data so that the customization layer for brands comes into play. And that's where at MIQ, we do a very exciting work in customization. We have a global data lake. We have 600 million consumers data lake in the India market and we are able to have a lot of third party data partnerships all put into a data lake through a data studio. And all this magic is of course done by our data science team. I look after the business part. But it's fascinating to see how we are able to enrich the first party data with our data lake and then figure out what is the right consumer segment for a brand. How do we take the planning and targeting to the next level? How do we look at the data analytics from a marketing perspective also to the next level? And what I also realize is that media is kind of becoming a bit of commodity and the data layer and the analytics when it is get attached with media, then a lot of value comes into it and that's where we are able to customize it for a lot of our clients. And we work with our agency partners very, very closely on that. We'll talk about a few case studies also, but yeah, this has been my big data to smart data journey. Thanks. I'm Dr. Ashish, just an asterisk. Please don't call me in emergencies. I'm just a PhD. Moving on, I'm working with Narayana Health and we are trying to shift from just being known as a hospital brand to an integrated health care brand. I think we've spent enough. Let's move on. Sure enough. So actually the first question goes to you, Ashish. So in the pharma space or health care, what do you feel is smart data? And how is it different from big data? Can you live without either of them or basically just open-ended? What does it mean to health care, essentially? What big data or to smart data is? So smart data is a definition we are still trying to understand and I think a lot of us. But from a big data point of view, I think it's one of the industry that actually works on personalization because each and every individual data is different. And there are more than 6,000 data points for your health care that needs to be managed or mapped over the course until the time you are 40 years old. So if you're talking about that kind of data points for almost, let's to start with 1.4 billion of us, that means it's huge, huge, huge volume that you are playing with. Now, from understanding or movement from big to smart, I think we were, while the big data came to the entire forefront, we were already applying a lot of academic filters by ourselves. We were saying, okay, fine, even if I'm seeing this, there can be certain other ways to look at it. And I think that is what transformed into looking the right cohorts. Not all 6,000 might be important for you given your age. Not all 6,000 might be applicable to you given your geography, given your gender, given your DNA. So from that point of view, you have to have to be smart in terms of knowing which are the ones that move in the direction with your age. So that's how we plan to kind of utilize it for the betterment and not bringing you to hospitals. Got it, got it. That's a pretty interesting case. We were just talking about how preventive is by itself a segment of its own and using smart data there is something that can trigger. Uteo, so what is your interest with big data and smart data? What is, how do you differentiate that in the FSI, the FSI space? So I'll take possibly one step back and I will try and draw an analogy in terms of what big data and smart data means to me at least. So every monsoon season, there is a fair amount of media coverage in terms of, you know, how much rain, what millimeters of rain the city has received, right? And then there is also coverage in terms of, you know, what is the percentage of the lake levels which supply water into Mumbai, right? So every day you have a news item saying Vaitarana is at 80% and, you know, Modak Sagar is at 60% et cetera, et cetera. Now, obviously what, finally we are trying to achieve or what finally we are interested in as consumers is the water which comes into our homes, right? Now, for me, the water, the data is the rainfall. The water which is getting accumulated in the lakes is big data. There is some amount of filtration of that data which is happening while that water is being transported to your house, right? Which maybe the municipality is doing or the water supplier is doing. Then once it reaches your house, you do another level of filtration through your water purifiers before that water comes to your glass and it becomes consumable for you, right? That is smart data. Something which is consumable, which helps me quench my thirst, which helps me get to what I want out of this entire, you know, universe of from rainwater to lakes to water supply to coming into your glass, right? And that's for me how I look at big data and smart data. Now, I would say smart data is possibly something which has been jargonized a little bit. It is something which inherently we have always been doing, right? The vast amount of data that we accumulate, all of us study that data, we work with that data to drive the outcomes that we are trying to drive for our respective businesses. In various ways, now maybe a lot of it was a lot of manual intervention, manual analytics that we were doing and we were doing all the analytical models. Now, you have a lot of algorithms which is helping you filter out the data and give you what is actually consumable to you. The reason this is absolutely required is also because the volume of data is so huge. There's two and a half, three quintillion bytes of data which is getting generated every single day in the world, right? Now, obviously we cannot store so much of data and storage costs are going up by the day. So how do we then define what are the data signals which we want to consume, which we want to retain, which we then want to filter and make it into actionable intelligence which will help us drive the outcomes that we want. So that's what big data and smart data is for. Nice, that's a fantastic analogy actually to put this in perspective as well. How, and said here, I would love to know how in the programmatic world and in media buying, like how is this, of course, we are crunching a lot of numbers. We are measuring our success with QPSs. How is it that the big data is more consumed and how does this make our smart data perspectives, another make our campaigns better or smarter? Can you just give your perspective? Yeah, we were discussing this, right? So we have a very interesting case study with luminous inverters. So it's a category where you don't get up in the morning and say that today I aspire to buy an inverter, right? It's a utility, but what you want to do is you want to reach out to the consumer when they are really thinking about the category and thinking about your brand also probably. So for this category, what we did was we found out a government website where there is a lot of data on scheduled power cuts across India, right? And so our engineering team created a crawler and got that data and fed it into the DSP, right? So this is an example of converting big data into smart data. So as an example, if in Jaipur, 2 to 6 p.m. on Sunday, there's a power cut scheduled. So the campaign will go live at 2 p.m. You will see the ad on your mobile during that time. And it is also a dynamic creative in terms of the languages vernacular. So it will be in Hindi. And geocontextual intelligence is also there. So it will also tell you the nearest address of Luminous Inverter Dealer where you can go and buy the Luminous Inverter. And for unscheduled power cuts, what happens is consumers will tweet, right? When I'm in my locality when power goes off and to express my frustration, probably I will tweet or so we collected all that data as well. So even for unscheduled power cuts, we were able to take the campaign in those areas in the right way. And so the campaign was a big success because consumers really got their information in the right context when they're really thinking about having a power inverter. 6x times benchmarks, huge amounts of footfalls, huge amounts of sales increase as well. And the agency with whom we did, they won MVs and a lot of other awards as well. So that's the power of converting big data into the right data, as you rightly said. And then I was able to really give the right information to the consumer at exactly the right time. Fair enough. Actually, the part that I resonate the most is that before we tried to solve, like getting into solving a big data or what AI or smart data essentially, I think the bigger ask is what is the core problem and what can data at times need not even solve the problem. But is there enough, again, see, anything with big data needs volumes of data to put this together. And then comes the ability to, if there is a volume of data, then taking that data, coming, making smarter insights off of it. But the core thing is identifying what the core problem is and then tracing it back. And here, we love to understand how EdTech is kind of taking this shape up. And what's the... You are taking me away from marketing a little bit, but then I think Siddharth will always do a better, justice to marketing than me, right? So I think the first premise that, which all of us understand is, if you go back in time, there was data, but there was not so much. And then we got ability to store a lot more data and we started storing a lot more data. So we became hungry for data. And in our hunger, we started collecting so much data that we weren't able to look at. When you collect so much of what you can't utilize is when you have a bit of a trouble, right? While, that is true, there was still, as mentioned by Mr. Day, that there was still folks who look at the data, create filters and try to make meaning out of that and then utilize it. Eventually, we've come to a time when, I think we've come to a time where all of us don't have patience, right? We don't want to watch a long-form video. We want to see a short-form video. And we don't want to spend hours running a report on SPSS and wait for something to come next morning. We want to look at some insight right away. So that's where I think all of us want to move to. I don't think all of us have moved there yet. We have a lot of data, but moving towards making meaning out of that data and using probably AI to try and look at the data and answer some of our questions is what, in my mind smart data, the term should try and look at. Now, the premise which you are talking about, right, is the most critical premise. Earlier, when you were trying to filter out, you had a question, you had a hypothesis, and you were looking at that hypothesis to find out and answer, that premise doesn't change, right? So if you want to have AI on top of your big data and it's not going to do anything unless you ask AI what the right question or what right hypothesis is. And then you will get a certain data as an output which is input for your thought process, and then you will have to still think through whether that makes sense to you or not, right? And I'll give you an example of marketing site first too. If you look at any media analysis tool and try to understand what your audience does, okay? And most often you say, hey, my audience is interested in entertainment. My audience is interested in music. But that's like given, that's like 95% of the internet population, right? That's just information for me and it's not smart, right? Now, I have to figure out what are the other things that the data is talking about and look at what's right thing to do, right? For EdTech, it's slightly away from marketing for me, but I think there's a lot of room for, you know, education providers, right? I'm not saying EdTech only, I'm saying education providers to try and look at the data that they have to identify how they can solve for some real problem, right? The real problem could be, hey, you know, my faculty has a very good rating, but, you know, the students are not sailing for example, you know, is there a drop rate in the lectures that they attend and things like that, right? Or the other way around, right? My faculty has a very poor rating, but everybody is stuck around to the lecture. Did he like not give them a very good score, right? And, you know, this is where he'll probably make, I mean, this few examples, right? But this is where he'll probably make sense for somebody to look at the data and identify what the real problem is, right? Because would you look at, you know, without putting intelligence, we'll probably drive you towards things, you know, which is absolutely bizarre. There enough. But in this market, right, which is, so in India specifically, again, we are now the largest in population as well. And the country is mostly focused on branding and identifying. Mostly the problems are towards customer acquisition, right? What is, and here, what should, why would businesses invest and in big data, actually, what is it? How does that enhance the end consumer experience through big data? Should the focus be, like, because if your core problem is customer acquisition and branding solves that to a large extent, but how has big data and AI helped in enhancing consumer experience and as a result, improved customer acquisition or retention for that matter? So is that, what's your take on that? Fantastic question. So, you know, when we speak about, so let's start with data, then arrive at big data and then try and answer the question. So, you know, each of us, all our organizations, we have our internal source of data, right? So, say for example, in a bank, we know the transaction data of each of our customers. And, you know, there are millions of transactions which are happening every single day, right? How do I make sense of those millions of transactions? Can I afford to sit through each and every transaction and figure out what is happening? Now, that is data which I have with me. That data will give me some sort of a sense of how Santosh is as a person, for example, right? But then the way Santosh interacts with a bank or uses his credit card does not define who Santosh is, right? So how is he interacting, say for example, on social media websites, right? Within social media websites, let's go, you know, one level further. Am I the same person when I am on Instagram or Twitter versus the same person when I'm on LinkedIn, right? So, there are a multitude of data sources which help me understand what Santosh is. What is he looking for? What are his, say, passions, interests, et cetera? And can I then utilize that data to figure out that Santosh is planning a vacation and therefore should I pitch and travel insurance to Santosh at this point in time? That's where you know that the journey from data to big data to assimilation to filtration and finally targeting the person at the right moment at the right time comes through. And that's where I think smart data comes in. I'll just pitch in. So rather than pitching him the travel insurance, pitch him how to make money for that travel, I think you will win him for life. So why you're on it actually. So here is again in the space of healthcare. Does it board up on to actually invest into big data and smart data? Do actually do customer acquisition or retention? How does that play in the perspective of? See, as a brand, we don't want you to come again. Absolutely not. And that's our philosophy as well and for the past couple of decades. And I think we're gonna be on the same path. But coming back to the question, I would not see it as acquisition or retention. Let's look at it as your journey of healthcare. We are currently working towards solving the problem of what next and giving you well in advance the things that can allow you to make it what next, what next. Just like how it works in Ecom, what's the best next offer? We changed it to by saying that what is the next negative that can come? But these are the four solutions that can definitely, if you follow, can definitely make those negative go away. And in a sense, how does this improve customer experience or in this, I mean, I don't wanna say patient experience, but how does this improve the customer experience? It's absolutely a consumer experience. But see, just like how he said that tomorrow morning when I get him, I'm not gonna talk about buying an inverter. And the whole experience of healthcare is under stress. You're not happy paying that money, right? Now, if you're taking out that stress out of that equation and if as a brand you're able to do that, then that means I think from our point of view, at least inside out, you've delivered the best consumer experience. And this will start getting understood by patients or consumers over the, say, couple of years or say a decade or so. But that's our intention. And that's definitely add to loyalty. I was gonna say advocacy of brand and the campaign essentially. I just want to quickly add here. So there's a different case study which we have, which is not consumer acquisition, but so there's a client of ours, cab player in US market. And so they were seeing this consumer behavior that some consumers would book a cab, but they will not turn up. So what we did was we basically understood the profile and we did a breakage analysis. And we said that, you know, okay, for marketing channels and even from an organic perspective also, let us not reach out to these consumers. Right, right. So it's the opposite of consumer acquisition. As a use case from there. True, true. And what I have also seen is in many of these cases, if you're able to harness the power of this, you can create the ideal customer journey and even devise those trigger points to actually make those who probably are in the same cohort, but have not taken that journey. And as a result, one is oppression use case. The others are identifying what the next best offer or next best action could be and take that up from there in. But may I tell you one thing? Like if a business needs to discuss or kind of get into this journey of, like should they explore a build? They build it on their own. And of course, I'm gonna get to you very soon, but because you've taken that journey through, but should they build versus buy? How should a business decide? When I say build versus buy, here is where, see, building the entire infrastructure of big data requires significant vision as opposed to, I mean, money's to do so, but here I'm asking what would be the priority index to actually build versus buy for business to take up this human best task of big data and AI? Yeah, Mike. Sorry, I wasn't hurry to try and tell you my thoughts. I forgot the mic, yeah. So, see, it's actually not build versus buy. Few things you'll have to buy, few things you'll have to build, and eventually, for example, if you want to store your data unless you are Amazon or Google or Facebook, you don't want to build, unless your data's so much, you don't want to build a solution to store your data. You lease it out from AWS, you lease it out from Azure, you lease it out from Google, and then anybody else, right? And then start your journey of storing the data. But that's about storing the data, big data, collection, right? You can probably buy some collection solutions, and that's fine, right? But when you want to get into the journey of collecting data to making meaning out of that, right? Or to turn that, there will be few off-shelf solutions which you can purchase, but in a lot of cases, you will not be able to get an off-shelf solution, right? Then you'll have to definitely go for a build now. Build doesn't necessarily mean that you build it yourself. You can hire a tech partner to build it for you. Right. But yeah, it's a mix of both that should work in most cases. Right. Unless you're very small and you know that there's something for your industry which is off-shelf ready, available from somewhere. And sure, Ashish, I mean, we were discussing about this, and what would be your chain of thought in this direction versus build was in buying this? The organization took our decision six, seven years back, and it was not that we had the resources or we wanted to build, and nor that we had a vision of how to utilize it. The requirement at that point in time was our current providers were getting out of hand in terms of servicing the kind of data that we had. Right. So when the decision was made, decision was made to just keep the ship running, the engine running. And today, I think it could be learning for somebody who's planning to do, but we can definitely say today, we are able to serve data. And now we are passing on that legacy and intelligence to other hospital chains as well. Yeah, and that's a great journey, actually. And what I have seen is at times when the decision, such as this has to be made, it also depends on how soon would you like to go to market. And when you do decide to go to market with an off-the-shelf product, then you need to know that it won't support all use cases. So we'll have to make a conscious call what use cases they are and shortlist because your priority is to go to market faster. I have seven minutes left and I have a lot more questions, but in interest of time, let's get into the rapid fire. So this, by the way, this question has no right answers. I have some seven of them here, and I think we should take about a minute. But here, please, whoever wants to answer to it, please do. And again, as I said, there is no right answer here. So. No hamper also. Sorry? Hamper. There are some, there are, well, I mean, we got to take collective responsibility towards it. So do you feel businesses are embarking on this journey of big data and AI, due to competition or peer pressure? Peer pressure. Competition. Competition. I think, so after COVID, a lot of brands or a lot of companies have realized that everything is going digital. And that's why they are now more aware of data and how they should use it, et cetera. So it's a mix of competition and, you know, digitalization due to COVID as well. Garret. Garret. I thought this was going to be funny. I think a lot of, was my answer will not be correct. It'll be a little funny. I think a lot of guys are getting into this journey so that they can waste the money that they have. It's that shiny guard that I want because my neighbor has it. That's true, that's true. Which industry do you feel has embraced this big data and AI effectively? I think Attic has done a lot, right? Then BFSI is doing a lot as well. I would say, Pharma, we are seeing a lot progressively. Yeah, e-commerce, of course. Garret. Yep, I think all the major players, right? Your Google's, Amazon's of the world. Hitech. Yeah. So they obviously are using it very well. BFSI, like you said, I was reading somewhere, you know, some global data survey which was done. So apparently some 23% of the overall big data sort of revenues are driven by BFSI companies. So to that extent, there's a huge amount of investment on this side of things for BFSI. And at the end of the day, you know, just maybe, I know there's rapid fire, but I'll take 15 seconds, maybe to go back to the previous point, you know, the shiny new car. I would potentially put it slightly differently. It's about whether the consumer is buying your car or your competitor's car. And for that, I need to understand the consumer better. So if I don't invest in understanding the consumer better and you know, like we've said at the top of the conversation, you know, the time to market is very critical. So somebody else will go ahead and sell their car ahead of you. True, true, sense. Which industry do you feel is not going to reap any benefit at this point in time? A very controversial and you may choose not to answer but also that's fine. Any industry that you feel is not going to reap benefit right now. I can't think of. This, sorry, you're going to say? I would say that for very large B2B sort of industries, their use cases are limited as opposed to B2C companies. But having said which everybody has need in varying degrees and that's why, you know, every company will have their own strategy. And like to varying degrees, there needs to be investment into smart data. Even during elections also, now a lot of big data will get converted to smart data and will be used. You get, you create videos now based upon how the consumption of that data is essentially. Absolutely true. How, when would you imagine this would have happened, right, if you thought of it? So do you roll your eyes if someone starts selling you use cases of big data or smart data? I'd rather the question is more, what facial expression would you make when someone is selling big data or smart data to you? So they roll your eyes or is it like? My reaction is curiosity, right? I would want to understand what they're doing, how they're doing it. Okay. Yeah. So there are more use cases still to be solved, essentially is what that is. Yeah, I mean, we are just starting. Got it. For me, I think what I am more interested in is in how other industries are using it. Are you using it? Fair enough. So whenever someone is wanting to have a discussion, I would be more interested, not in terms of what they are doing for BFSI, but what they are doing for SEA healthcare or SEA net tech. Fair enough, fair enough. I think I'm always curious to know. No. Got it. Always. Do you feel we have enough talent pool in India to manage the big data? What do you feel about it? Is it fantastic, good or bad? I think just throw any problem at us. We're always good at it. Good at solving it. Yeah. We are good, we will be fantastic for sure. Fair enough. Any take? I think there's a lot of talent in India. What I see is that we should have more and more apprentice kind of programs where students and apprentices are able to work with the industry. And really, because whatever I've learned, whatever I've done, I've learned through my job. All right? So I think that's an urgent need for us to do it for sure. Got it, got it, got it. Another? I think we have shortage of talent. I feel that too. I think it operates in cohorts where some parts we are really good at. And if you take specific technologies and look for talent in it, there is shortage. And the shortage has driven up whatever their asks are as well. So it is becoming a tough spot to be in, essentially. I mean, while I'm saying that, I'll not say that we are not producing enough. Yeah. But we are exporting a lot more than we should. A lot more than we should. That's true. A fun one here. Between Darius III and Alexander, the great, or would he still be the great if he had access to big data, if they had access to big data or AI? Who would have won the war? So I'll rephrase this. Between Darius III and Alexander, who would have won the war if they had access to big data and AI? This is 329 BC, so. No, if that time there would have been access to data and AI, probably a peacemaker would have prevailed. That's true. What a word. We needed them to also have, you know, Instagram and TikTok. Then there was no chance of no war. If I took a very controversial stand, I think Alexander would have still won because he was actually in it. He was leading from the front, so chances are he would have been thinking about what the vision would be towards winning the war and then trying to kind of use big data the way it is. No? You're already, I already know what your next move is. Right. Which department in the company would you completely offload to GPT and make that as a primary touch point? It could be internal, external, but what department would you completely offload to GPT? Again. I'm not sure. Please feel free to pass. That's fine too. I'm not sure if completely, but I think a fair amount of the MIS work, which is done, can be done, financial analytics or got it. A fair amount of, I would say, dashboarding, for example, and a lot of the MIS is that the team spent so many hours to create, or now we are seeing some use cases in terms of PPTs getting generated at that click, how good or bad they are. That's a discussion for another day. Just add on top of it all the mundane jobs. Because you're going to see those excels day in, day out, you'll see those PPTs day in, day out, everything, like which I will not come to office for. Right. Let's finish them off. Repeatable tasks, essentially, that's true. This is the last set of questions. We are two minutes over time. I'm sorry. Do I still have time? Oh, man, she's okay. So let's take those questions then. All right, some special questions coming from the audience, I believe. We'll take, I mean, there are many here. So I'll try my best to answer these offline. The question is, how to use the data that helps us to navigate and optimize to increase the latency with the help of JNAI or follow the compliance? As follow the compliance, okay. So basically, I think to interpret this would be what would help optimize and navigate using JNAI? What kind of data helps to do that? Look, so JNAI is something which is still, I would say, a fair amount in its nasancy. The problem, so while JNAI can help solve for a lot of things, and we are, every day we come across new use cases where it is getting used, the problem that AI or JNAI will always have is that we do not know what has gone into delivering the outcome. You will only see the outcome, right? So especially where we are looking at analytics, what sort of understanding has gone into processing the data? What are the data signals which have been captured and on what basis has that outcome been given to you? Is not known to us as humans. That is not something that you get from the machine. And that for me is something which we need to carefully consider. So if we go into the market, just based on the responses that we get there, there lies a risk of it also not working at some points in time. Yeah, the hallucinations or the artifacts of JNAI, essentially. There's another question here. How will smart data slash AI and big data in collaboration impact media and advertising ecosystem? So would you want to take that? No, I think it's already impacting a lot, right? Meaning all the big tech are using AI extensively. And so the more the data comes in in terms of digital format, it can be used to understand the right consumer to be targeted, right creative to be communicated. And also in analysis that, you know, which are the segments which are really working well for you. So already so much of usage of AI for media is happening. Can I add something? So we are in a world where we see probably about 150 to 200 ads a day on an average for that 650, 700 million Indians that we are. And probably some of those who consume internet so much more probably see a lot more ads than they should. We are also getting into a situation where a lot of us don't want to see so many ads and we have blind spot for those ads. If AI with big data can help a market here reach out to the right guy, minimize the waste, it essentially means that I am not going to sell so many ads and still reach out to the right guy. I am ready to pay the money that I was going to pay for those three, four, five additional impressions instead of reaching out to that one guy. To that one impression, which essentially mean that industry at some point of time will be able to reduce the collision of ads, reduce the frequency of ad and increase the yield, and improve the user experience. This is possible. I'm not saying that this is going to happen. But this is a very, very strong use case for marketing platforms to look at. While I'm saying that it's not going to happen tomorrow. It's going to take a lot longer. But yeah, a lot of investment will have to go from tech platforms for that. And I will to say that I will reduce the number of impressions. I have many more questions here, but sorry, apologies. We are six minutes over time. Thank you so much, panel. It was an amazing round with fantastic experience from there. Thank you so much.