Don't Build a 100% MarTech Stack: Birlasoft's Satinder Juneja on AI, Agents & B2B Marketing

Hello and welcome viewers. I'm bridge pawa and today I have with me another amazing Indian leader Mr. Satendar Junja. He's the global marketing head of billoft. Many of you would know him. You would have seen him speaking in various conferences on marketing pan India. Thank you Mr. Junja for uh giving us your time. >> Thanks bridge and thank you for having me here. So uh you know let me dive straight into you know and set the context and we were having this conversation before the interview has started as well. Um about 15,500 plus tools in the Martic ecosystem and um every year they are increased. So last year [clears throat] last year there were about 14,900 um so every year more tools are being added. How difficult has be it become for a uh for a CMO or a head of marketing to actually create a stack today and what sort of architecture they are using? Are they going into integrated? Are they going into composible? Can you tell this to us from a CMO's POV? >> Okay. And Okay. So normally when you start the interviews these are you get easy questions but you've straight away you know got to the core bridge. Thank you for that. Look uh this part of how do you pick the stack? What is the ideal stack? How do you create the stack is probably a daily worry of of a CMO or a marketing leader and it is only getting complex. Right as we speaking, we were speaking before this 15,500 or 600 number has swelled from 110th to less than you know 10 years. It's kind of swelled a lot >> and um but that is about martekch and that is the beauty of any technology. >> The the question is not about the technology per se. So that definitely I mean technology does whole bunch of um easy things for you making things easy for you. uh but be beyond besides technology there's another important question of um how are you running it now when you are thinking of a stack you have to look at [clears throat] both things in the same uh on the same plane right technology will do its work whatever it is designated for but how are you planning to execute and how are you planning to manage that is equally important now if I put all of these two together now if I look at how do I put my stack together, right? Is it integrated? Is it composable? Whatever it is, my view is that it has to be composable because your new demands are coming in. Now, if me and you were having this conversation in 21 year 2021 or 2019, you know, AI was not to be thought through. Probably the max most complicated thing in 2019 was intent signals, intent data, right? some amount of predictive analytics but this was again based on data and data modeling. We were not talking about AI. We were not talking about agents taking over almost complete roles in the marketing you know team almost complete roles being taken away. So we were not talking like that. That means if somebody thought about a stack in 2019 and or 21 integrating AI would have been extremely difficult now if it was not composible. Right? So that's that's my you know rationale of you got to create something composible and I think one lesson that I've learned personally after AI coming in and coming in with such a force such a such a velocity which is direction and the speed to both put together is don't create 100% stack ever >> create probably 85 89 90% leave 10% room Now where would you leave 10% room? I don't know. I mean would I put would I put it in the content layer, identification layer, CDP layer, uh privacy layer, I don't know. But somewhere 10% flexibility. Now it's not about leaving it incomplete. It's about building in for 10% of flexibility at any given point of time so that you can quickly shift the the the the way forward. Right? because business would need it. Your audiences will grow and go to whatever is is needed. Your stack has to responding to that and you as a marketeer need to respond to that stack right you are the one who's responsible for playing putting it together running it and when you as a marketeer look that is where how it is coming back you know from overall technology to your customer from your customer to the stack stack to you. So [clears throat] you are in a way last in the food chain but in the most decisive manner at that point of time you just can't look at technology you have to look at uh you know how are you executing it that brings me to the next point >> I don't know if you want to ask it but let me just give it to you see putting technology stack and a complicated one is the easier parts frankly speaking but running it >> and planning for it to run run it is equally difficult and that is where uh your your ability to force see the future or your ability to foresee the situations wedded in your context, wedded in your reality, which is your organizational reality >> is important, right? >> And how would you do that? I think the answer to that lies in marketing operations. So, while I may get swayed by the glamorous of shiny tools of mic tools, right? But in order to run that properly, you know, your operations have to be and span, right? And that is where people I mean another question is human inside versus AI lot of >> that was what my followup would have been to this >> right I think when you're talking about operations that is where human and agents start to work beautifully >> right so very interesting because you you you say that if we would have been talking five years ago you would have had recommended an integrated stack because uh you know AI was not there right but now you would recommend a composible because it makes it easier for you know for marketers to have that layer of artificial intelligence on top of it and make sense of what is coming in >> right >> but but then you talked about marketing operations now when you say operations right now most things that we are doing currently are now getting automated especially from the content POE right so where is is that marketing skill most valuable today in terms of operations right if any young marketer is looking is watching this video, what skill should they come in with? And when you say operations, uh if you can elaborate on, you know, human plus in addition with agents, how how does a how does an organization's marketing operation ideally should look like today? >> Okay. So, so when you look at marketing, you know, or I mean what's so you two different two questions, right? What should be the skill set of a youngster, >> marketer? >> Yeah. And how should operations be looking like? First one which is the skill right. Uh 10 years 15 years back the conversation the marketing was is marketing a science and art. 20 25 years back and by the way I come from last century right? um 25 30 years back it was more of art you know the creative the design and things like that then technology started to seep in right uh as the technology started to seep in and the question was marketing science and art and it was more veering towards science I remember speaking in one of the conferences some 10 12 years back and I said look I don't know which 50% but it is 50/50 science and artill said 50% of my budgets work 50 don't I don't know which one somebody said that I am here the case that whether it is science or art I don't know but a marketeer has to be the mix of both. Today in 2026 a marketeer who doesn't understand technology would not be a successful marketeer. So understanding technology is par for course. It's like it's like knowing English or knowing the language of the business. Right? when I was getting into the workforce rather at a young age people would tell us look knowing powerpoint word and excel is very very important >> and then it became part four course so today you don't ask anybody do you know computers it's done [clears throat] a 10th class kid would knowing would be knowing the so knowing technology and knowing what technology does in marketing is part for course you can't escape from that however the human insight the ability ility to assess things. Observe. Now, observability is becoming even more important. >> Human observability is becoming even more important. Okay, we can we'll unpack it >> during the course of the conversation. But your observability, your curiosity is very very important. Your ability to assimilate diverse signals quickly is very important as I said. So for example one example I'll give you a bridge in last 6 months how much have you heard about prompt prompt and prompt engineering >> hardly hardly >> right what go back one one and a half years back only conversation that you were hearing about AI was prompt prompt engineering prompt engineering and all that right today you're talking about autonomous agent take and all that right and tomorrow it'll be something else what I'm coming to that is the pace of narrative and pace of requirement is changing very very Imagine if you are a youngster 3-ear work x in in in marketing >> you start when you when you walked into the marketing but when you took up the job in 23 AI was just starting >> you said okay let me just look at chat GPD and some prompting and now boss you're talking about agent take and all that you know there's already second or third generation of AI that is coming in now what is the kind of skill set that is needed here Right? That is so your ability, your curiosity, your observability. Another important point and I'm digressing a bit bridge and forgive me for that is the pathways to expertise are collapsing. I'll repeat the pathways to expertise are collapsing in the current age in the age of AI. For example, you you in my world also where I started there was a hierarchy. You spent 3 years here, 5 years here, 7 years here and then you became GM marketing and then you became this and you became that. There was a certain amount of knowledge and skill that you had >> gained through the years of experience. Today all of that can be done in matter of days forget weeks right but what you're not getting is the observability what I can observe after nearly three decades of experience and I'm not kind of demeaning anybody but probably somebody with five years experience don't has does not have that ability and insights observability and insights so >> the wisdom of age >> sorry >> the wisdom of age >> the the not only wisdom of age It's it's the sheer cycles that you have gone through. See you can today put whether it is in Claude Kim or Chad GBT what is the role of CM you can get it right but the cycles of failing and succeeding and what that rub off on you cannot be >> repl so so far AI can't give it to you maybe 2 3 years down the line yesterday my young son was telling me that they have started to buy books and they're scanning the books some AI models and AI engines right and then they are shredding it >> and etc etc Maybe maybe 3 four years later even that wisdom would probably get replicated but not today. So you ask about what is the skill set. So it's a mix of this skill set that people that youngsters need to know. Theory still is not going anywhere. We still are wedded to four Ps of marketing that Kotler gave us some 40 50 years back >> right um but the things have got added on to it. So this is about the skill set of a of a marketeer. I'm not getting into specific certification and courses. >> Yes. Absolutely. >> Right. But the overall temperament the learning temperament learning mindset the second it's not about youngster I am I'm honestly telling you a bridge and you know over the last weekend I've gone through a massive FOMO on learning because one of my friends few years junior to me in work experience sitting in Singapore has created a Bhagwat Gita's app equivalent and you ask and this is his weekend VIP coding project. I mean he follows Bhagat Gita that is fine and he gave it to me just test it out. I asked three questions very philosophical very you know cross-linking and all that and the it gave me phenomenal answers and I'm like if he can create it where am I for and for marketeer biggest and I speak for marketeers only right the biggest need is to stay relevant moment I see anything that will make me irrelevant I get FOMO so I'm like for last week or so so I have not done VIP coding you know what will happen. I'm seriously going through that. So the the learning mindset is just not for a youngster. It is now a continuous thing. You continuously will need to to learn, come up, pick up with new things, experiment, fail, move ahead, whatever is the typical learning cycle that is. >> So that is the that's how you need to look at learning. I'm sorry I kind of extended it beyond what what you wanted. Now coming back to inner marketing team what the operations would be when we are talking about automation or agentic world here is my question back who's creating these agents you are for your need for your context so agents and human I'm I'm sure this is oft repeated will exist together here is the deal I'll again go back to illustrate certain example at one point of time payroll were done by hand. One of the first things that it did was to automate that payroll when computerization happened. Then it moved to bank accounts and things like that. Right now when when that happened um then the Excel progressed then Excel came with this macros, Excel came with visual basic programming right all of that only helped accelerate. So those were the agents of that time automation agents. First it was basic spreadsheet Lotus 123. Came Excel with graphics UI. Then came the uh the the macros in the and the uh coding you know visual basic coding within Excel. >> But all of that was trying to solve the problem of data data analytics record keeping and you know insights out of that and representation of that. Same is the case with the agents. So today you have minor agent. I'm just making it up. There's nothing called a minor agent but I'm just making it up right now. Tomorrow you'll have a major agent etc. But they'll have to coexist with human being. It will only make humans or the overall output sharper, better, faster. >> Right? Uh now how do you imagine the marketing operations? So first you have to uh you you need to have your operations in order. I said this is where we started with this conversation about Martekch. So once you have your operations in order then you'll have to figure out this part can be automated reasonably with with reasonable predictability with lesser risk that is what first you will uh automate and then you will go on to the more riskier riskier stuff because >> uh your comfort would also go up >> with the technology and the behavior of the technology. It's just like learning the car. when you learn the car you know you you don't even go to your fifth gear or fourth gear for first 2 3 months >> but once you start to go there then your confidence builds up and then someday you are zooming past even on a smaller road I don't recommend speed driving but you are able to drive and navigate easily forget the the highways right so so your confidence will also grow once you work with minor agents then you'll create the major agents and major automations will come in so but then you will have to imagine what is the task that an agent will do which will be different in Basoft marketing team versus let us say I'm just making one name Coca-Cola marketing team versus butter marketing team I'm just taking these name from the different industries perspective right so this will be different so so so marketing team members in Berasoft including me will have to imagine what is and where do we need the agent with what frequency and with what autonomy Right. [clears throat] So I think it's a it's a these are interesting times for us to see how this will move and how this will evolve. >> I don't know if I've answered your question or >> No, you have. I have. My next question actually follows up and know for a fact that you work with a lot of companies, you know, as marketing head of Pillarsoft. Um what level of agent penetration have you seen? So I'm now going to the you know the followup which is what sort of examples you have seen because you know we have heard stories about an airline using AI agent systems and customers complaining you know and then somebody successfully using it but but but from an inside POV like working with these companies u you know what have you seen how far has that penetration reached and I'm I'm completely on board with the fact you know that Times are changing very fast, right? So we were talking about prompt engineering, we are now talking about agents, right? Um and maybe two months from now there will be a different conversation or 6 months from now there will be different conversations. But as far as AI agents is concerned which is the the conversation right now what what sort of penetraability are we seeing right now? My my understanding is talking to some of the peer levels and observing uh with in in my industry and outside what is happening practically on the ground we are take we are reporting is getting identified a lot uh summarization checks black checks duplication checks >> happening uh a use case I'm telling Yes. >> See you we use uh stock images from the libraries and the biggest problem for any marketeer comes is when somebody else has used the similar stock image. How would you know? You have a library you pay whatever $15 20$20 $100 to get that image but you don't know uh let us say bridge you are my competitor right? Uh how would I know bridge has also gone and downloaded the same image. I have no way till the time you publish it. So I been discussing that we need to build up an internal agent which can go and crawl some other websites through external observability and see if same images has been used. Right? So one is the reporting summarization um and you know follow-ups are becoming agentified right that is also happening and checks that is these I give one use case in a checks but there are much more checks that are happening >> record keeping all of that is on ground now practical identification that is happening right >> automation is some of the evolved marketing teams where the experiments are running are bidding for SEO or uh you know media buying signals >> getting you know uh quickly on your screen is also getting agentified right you go >> and in fact for B2C companies a lot of personalization is also becoming agentic because uh for B2C D2C that scale of personalization which was not possible earlier. I think is really helping that because >> Yes. Yeah. So all your even your uh these quickcommerce sites and all that. So they're using a whole lot lot of agentification to personalize the experiences buying buying cycles etc etc. But my view on to that is pretty limited. I'm more on the >> I know I know that yes it side. So I'm telling you on ground this is what is actually getting agentified. uh two different functions linked with an agent is in the pipeline. some experiments are running. uh is that going to be um popular in 2 months or 4 months is anybody's guess but I think that is where we are moving as a marketeers towards the next time right >> uh so those are the agenty experience uh stuff that I'm seeing and then um people are learning how to create end to end agents right from the brief to the campaign in the box kind of stuff right so those teachings are happening. Have I seen a successful implementation of of that? I have not. >> But is that in is that in cooking? 101% that is happening. Is that 2 months away? As I said 4 months away, >> anybody's guess but around the corner, right? And I was telling you about the reporting and everything. M >> I am also seeing some amount of uh intelligence the insight intelligence also getting automated and coming to you more frequently right thanks to agentification okay so those are the some of the places where I've seen AI agents in operations and planning >> playing a role >> interesting So let me let me take this conversation um uh to discoverability, right? >> Yeah. >> Uh you know marketers would, you know, put money into SEO rankings, right? Uh being number one on Google, right? And then the other marketer would put an ad and if you search somebody's name, the other person's the other company's name would come up. That has completely changed now, right? >> Yeah. a user is now going to uh you know and I I was interviewing the head of marketing of Acer last um week and um I I did a a real time experiment. I searched for best laptops from 50 to 60,000 on charge GPT which most people are doing nowadays right and it gave me five options. Now if your brand is excluded from those first five options, you're not going into the let's say a Chroma store or a Reliance store to even check that brand out. You are just comparing the first five, right? Now Acer would have had invested a lot in the last decade in Google, right? But now the the dynamic is changing, right? uh now you have to invest more in you know geos of the world and you have to uh become the first option in searchability discoverability searchability right how difficult that is that that makes life of a marketer now because the you know with all of this coming in the whole as you said right the FOMO part right uh the tools are changing the paradigms are shifting right searchability has now gone from SEO to go How how difficult does that make a life of a marketer? >> If I may use Hindi words. >> Okay. Okay. Jokes apart. Look, technology was one way or was always easing your life on the second side was always posing the challenges that you hadn't planned for. >> No, but this is disruption. This is not a mere technological change, right? This is this is you have learned something of the last decade. Google rank and suddenly now it's not that game anymore. 80% are people 80% people are going to LLMs and searching for information which they would search on Google earlier right that's disruption that's complete disruption >> you are you are open to to contradict it a bit bridge >> I'm a journalist [laughter] and I'm not a journalist who is doing a TV debate so very very open to yes >> okay uh take the technology aside okay uh whether Whether it is AI or normal uh internet the need for getting discovered and ranked ahead of the others is just the same. This is exactly the same thing when I started my career. I started in print by the way I started my I was a management training in Hindustan Times. >> Oh amazing. Yes. The mo what was the most important property and Hindustan Times was number one newspaper of the country at that time >> was the front page solus front page quarter page ad that comes it used to be known as front page solus that was the most you know priced property and by the way we sellers we marketers will not get any credit for selling the front page because it sells automatically you're not making any effort there >> absolutely from there to TV spots. TV sports I mean let's look at cricket match right and then people started to find you know innovative places to put the spots the need is still the same bridge right comes internet comes you know people are crawling internet for their uh uh you know answers okay that is where I need to be number one and comes Google and they make ton of money ton of money ton and ton of money on on the back of that now comes AI the need is just the I need to be shown there first. You're talking of top five. If you ask any marketeer worth the salt, well, I have to be in top one. Forget it. I don't want to be number two also. Right now, the need is just the same. Technology is only the vehicle. Your ability to understand the technology and maneuver through that is important. 15, 18, 18, 20 years back, there were similar questions about Google. How are they ranking? We don't know. Then there came red hat, black hat techniques, keyword stuffing, hummingbird update and all that and all that came right and if we just go back 3 four years people still had not you know completely unlock the Google's algorithm >> that keeps changing yes anyway >> changing right and Google I mean Google is just one name but they also kept everybody on the toes because if people discover then you know >> doesn't make sense Google. >> Same is the case with AI. I have had pitches from the vendors. I had I have spoken to certain experts. Nobody's able to give me a firm answer that do do these three, four, five or 30 things and you will start to rank in AI or AO or G >> discoverability, >> right? The discoverability is still not there. However, however, one thing is constant and I'm still going to, you know, the the internet era and AI era on the discoverability part. Your brand's trust and credibility will make it show up. >> H >> I'll tell you and this is a question to me also from my seniors. Hey, what are we doing about it? How are we ranking and all that? We've had a one one case where somebody put us in the uh for a particular set of technology said who are the top five in this set of technology on 2nd October. So 2nd October the the gentleman did that got our name 3rd October wrote the email fifth or 6th October got a response from us by 29th 28th October he had signed uh you know non-disclosure agreement this is that's a first milestone in our industry that you are having a relevant conversation so for me so the way in my industry it works is you work you get into a long list short list and then you fight and win right so you started point is that if I'm not on chat GPT and now with Claude, I'm not in that short list. If I'm not in that short list, I don't stand a winning chance or fighting chance, right? To to to to win. So now what that gentleman saw on 2nd October was the great content that we had written some three and a half years back but it was highly highly relevant right for the technology where we are masters in right I think uh you'll have to find avenues and now suddenly in AI era I'm reading that Reddit is becoming very very important when we were starting with with uh Google era or or Google discoverability. Uh at one point of time blogs was very important. >> Blogs were yes. >> Suddenly people took this and marketers like me not people it was folks like me only. Oh b 30 blogs in 30 days. Anybody who can write a blog gave it to me. Then came the mobile optimized pages. >> Then came the responsive website. So so these shifts you know marketing guys getting shepherded in by these technology shifts are common. And that is why bridge I started with my conversation that this is not a disruption the way you are the way everybody see and even if you let me just go and unpeel this on a slightly technical note at least from where what I have heard people who claim that they will get us ranking high on AIO and go the basics still remain the same crawlerability is still important right your site map your site structuring is still important probably what AI is doing is earlier you were just concerned with your own website how is it structured and all that AI is now linking it to I'm just make taking one name which is Reddit here maybe some other forum here maybe some other and probably is now gathering a score and then putting you up in top five or top 10 now suddenly everybody will run to create forums on Reddit and things like that I mean as I heard we all do that it'll continue to happen but the basic steps still not change uh bridge if people you if your trust and credibility is built up the show show will happen you'll have to make an effort that is for sure you'll have to increase the crawlability you have to show up I'm just making it up again is on Reddit let us say tomorrow there is another forum and that is also getting highly referred by AI and Jio and all that you will have to go and make yourself present there but the basics of your trust credibility and differentiation are become even more paramount Right. Interesting. Can can I share an interesting uh like uh just yesterday I was reading in fact uh you know with diverse LLMs the answers are diverse right the suggestions are diverse so you never know which ranking you come in right and there is only a 6% overlap between answers between claude Gemini GPT etc and there was a study done and I was just reading about it yesterday and it was very interesting because you know the difficulty a marketer has to face >> while while you're very right that the basics don't change the foundations don't change but that constant need to evolve now I think the urgency of it is more than it was ever in the history of marketing >> this is what we spoke in the earlier part of the conversation as a youngster what is the skill set that you need to have so let us say you are a not even a mid mid-level marketing manager you're just below mid-level marketing manager would you not be grappling with this problem see 8 months back nobody had heard about clot Yeah, >> it was only about chat GPT. Now today people are swearing by claude right in India we don't hear about Chinese uh LLMs a lot but >> deepseek now beyond deepseek there is Kimmy also that is coming >> yes >> I mean what will you optimize for see in Google era it was only >> it was only one problem that you had to fix right Google and that competitive bidding which you alluded to earlier right I could I could put a bid on Bridge Pawa's name and if I'm paying more you know my website will show up thankfully you know that got quashed in one of the there was a judgment against that uh few few months ago >> right but that was the problem that I was grappling with now I need to think about chat GPT I have to think about Claude and you know I don't know who's the third guy coming up today and as they say catch them young if there is some other you know GPT or LLM brewing up right now in in a younger stage which could become a a giant tomorrow I better be there too right so people who have who would have thought about claude 8 n months back would be an advantage today right because they are beating the herd so I also have to look at the smaller GPTs or LLMs which could be becoming the giants of tomorrow so which is where again some amount of flexibility in your budgets and the efforts have to be always there sorry you were saying something >> sorry I was just adding to what you said earlier that you know you if if I think from a marketer's lens I mean um I'd rather invest in trust and credibility because even if you are going for a laptop search to any platform it will show you Apple as first right >> so I think I think completely agree with your point that even while there is an urgency to evolve the foundation of the the the basic tenets of trust credibility putting out the right content getting the right customer feedback back is still as important. It's just that the channels have changed. >> Exactly. >> Mediums have changed. >> Exactly. Exactly. That is that is precisely on the on the on the dot. So yeah. So as I said these are interesting times but I think how do you react to it? What kind of flexibility you build into reaction >> will will determine the the tomorrow. >> Interesting. Interesting. Uh uh coming to the question of data >> right >> uh now uh many marketers still I mean most marketers still grapple with uh data as a set and uh making sense of it even when you know we are at a very advanced stage of uh uh you know stacks and creating stacks and working with them CDPs layer after layer uh in fact I interviewed David Rab he's the head of CDP institute uh one of the earlier proponents of customer data platforms >> again luminary. Yeah, >> absolutely. Uh data still remains a huge problem. So I I one example I generally use is that you know I was once scrolling my Instagram and I was given an advertisement of in between two stories I an advertisement of a 10 lakh Mari Suzuki car came and after five stories an advertisement of a 70 lakh rupees BMW came right for the same customer which is me right so one of the marketers has gotten it wrong which bracket do I have not in they have put in money into personalization they have put in money into ad targeting and yet they are getting it wrong, right? And I think that is the same case with whether whether we talk about B2B or B2C, right? Data always has been a huge issue and now with added layers of stack and artificial intelligence. Um it was thought two years ago that data won't be an issue but it still remains an issue and the the example I'm talking about is very recent about a month ago, two months ago. >> Yeah. >> Right. So what's that we are still as marketers not getting right about data and how to manage data. I think uh I can't comment on bridge that what are we not getting right mistakes everybody does and maybe the somebody was experimenting with some signal which didn't land well in your feed >> right um but I'll go back to my earlier point of marketing operations you need to stack your data properly right the prep of that is important as you uh plan for various activities one of the one of the strong use cases and this is thanks to AI only AI and the mix of AI and data right put together is the evolution of synthetic data where we have experienced some success okay a positive use case right so let me just lay out the use case to you see what happens is in in my industry the buying cycles could be long it could be as long as 9 months okay any deal from prospect to deal landing in IT services for 6 months is a great stuff right and if you do it in 3 months you will get probably gold platinum silver you know diamond medal for that but this is how the nature is now what happens is behind that like just like any B2B there is buying center beneath that right and you will speak to the complete buying center different roles uh to get your goods sold first pitched and sold As a marketeer, I'm always trying to pitch to those buying centers. Now, if you look at any IT services firm or any B2B, the number of customers is very less some companies and I know some large B2B companies work with maybe 100 to 150 customers. That's it. And we are talking of billions of dollars. That means there is a buying center. There are various Google studies on what a buying center composes of. It could be anything between seven members to 12 members. >> At any given point of time, I am chasing,000 leads for example. That means I am just talking to 12,000 people. Those are the people I need to at any given point of time. That list will churn and all that is a different matter. Then again I'm talking of IT and I'm talking of B2B industry. Now for me to get access to these prospects is extremely extremely difficult while I want to sell to bridge power but he's still not my customer and how do I get to know what bridge will react to my ad like I have two ads one has a human face and second is a robot I'm just making it up right now let us say bridge doesn't like robots he likes humans right for me to assess that is very very important because then I will be putting that human ad or robotic ad in his feed. Right now I don't get that insight. So here is what we did in Bitlasso. We use external observability signals, externally visible signals to figure out what was a typical role in a buying center behaving like what were the information points? What were the points on which they were reacting externally observable? Right? So for example, bridge your LinkedIn feed is open to the world. >> Let us say you've always liked all the cricket related stuff. You've always commented or liked or whatever. It's easy for me to understand that bridge is positively disposed towards cricket. [clears throat] And this is one signal I picked it up. Similarly, with help of AI, we picked up maybe two or one and a half to two dozen signals for a particular profile. does and then we replicated it fast because this is data and this is AI replication is easier and faster. We were able to create a synthetic profile of various roles in a buying center. >> Once we did that now here is now the human validation comes my marketeteers work for these profiles day in and day out targeting them so they know where they are successful not successful etc. So first validation was done by the we call them business marketeers was done by the business marketeers. Okay. Having done that further improvisations once we marketeers felt that we are like 60 70% there we went to the on ground salesperson who's living and dying by these buying center rules and we said boss please can you tell us if this is more or less there or not? How will you you know give us a feedback and all that? is a it's 80% 70% and things like that. Moment we got a 70% kind of a cut off we started to put that profile in our project and we started to pressure test our campaign and reaction with this synthetic profile. Now you imagine that me getting access to bridge power was so difficult. A he would not entertain marketeer. He does entertain sales guys but he's not client of that sales guy. Still very difficult right? So as a marketer I have to go to the sales guy through the sales guys I will go to bridge power right now I cannot but through external observability I picked it up and then I replicate it right again because of AI and data replication is not I'm not saying easy but it's not strainous right and once you build that then the model keeps learning and you keep feeding the more signals and more observability that you've seen >> and the more robust the model becomes. So for us this synthetic data project has been an eye openener in terms of the cap the the potential of you you you mentioned about the marriage of data and AI >> for us this has been a a great learning eye openener and a bit of a success I would say >> uh is is this synthetic set >> that that we've been able to do that now we definitely are looking for bigger and more bolder experiments with this >> very interesting uh I mean still trying to process but you know looking ahead now uh with last part of the interview >> uh what do you think will define the world's best B2B tech marketing organizations and uh and what have you learned from the accentures of the world uh you know do and what can Indian marketers like you working in enterprise B2B the IT tech B2B deploy. So I think uh going forward I think and learning as I said is not only from one particular accenture they they do some phenomenal work but learning is across the board bridge right one is learning from different in fact I'm more keen to to learn from different industry what a D2C or a B2C brand would be doing than [clears throat] IT services no no offense or disrespect to the great IT uh services hands but uh what they are doing and can I import some of that or you know but I think going forward couple of things uh will become very very important one is uh you'll have to balance automation with human creativity that is supreme look there was this automation I just gave you the use case of synthetic data but somebody one of us got creative hey listen we are getting this we are getting this we are getting this so how about putting this And let me tell you bridge I'm just breaking it in a between it all started with when we were evaluating the intent signal for a particular services right so there is this weightage that you get on the Google keywords adwords and all that and we said okay this is happening this is also coming to us and this is also coming to us can we put all these three together and then see can we come up with something like that that gave us a chu success which we build on and which Then I told you the complete project that came about the uh synthetic data. So I think it was if I didn't have that automation if I didn't have that signal availability the creative would not creativity would not have been able to do the job. I would not have been able to connect the dots. Right? So my creativity lies in connecting disparate dots and creating a different shape altogether. Right now those dots is the that automation and the technology that gives you but you may be linking four dots to come up with different shape. I'm coming up with a square and somebody will come up with just dot will come up with a probably 3D square. Right? That is where the human creativity the perspectives uh that that matter. So I think that ain't uh going going anywhere. Uh second is uh we are already getting into uh context engineering that is another experiment that is at a experimental stage. Uh so that I think will go go beyond typical GNAI you know stuff in in marketing. Second is we >> can if you can elaborate I know we are coming towards the end but I would want to learn more about context engineering. So you have to teach the AI about the context, >> right? >> Uh how do you it's more or less like you did the prompt engineering, right? And and it is the learning time for us also. But how do you pick the context of your client's possible situation where you are not there and can you marry that to your context to come out with a situation? So for example uh one context is that the client is at a trade show. >> I'm simpl oversimplifying. >> Yeah. >> Second context is client is reading your email. Right now these two contexts are different. Objective is still the same. You want to be the favorable brand in both the context. So you will act differently. Right? Uh now this is a physical or digital situation. If you put that in AI, you will have to look at how do you engineer the context to client situations differently so that your ability to target the client using AI improves. >> Right? Second stuff that uh I am reading about it I don't know I'm just putting a wager on that is uh neuro uh neuroadaptive design probably or neuroadaptive [clears throat] experience because as the AI advances and the see digital is advancing AI is advancing your ability to connect various data points and bring that to experience layer only goes Now you spoke about high personalization. Now can you imagine what is the next in the high personalization that for 4 days you know uh Satendar asked for uh milk at between 7 and 8. Easy prediction 50day >> but what beyond that? What beyond that? So somebody needs to imagine and you know plan for it today. [clears throat] So I think uh AI will as I said if if this fractifies u which is neuroaptive uh design or neuro adaptive experience design if this fractifies AI probably will be able to read the the the emotions uh in in real time and adapt your experience accordingly. Right? So today Apple watch is getting lot of data from me today. I may be, you know, one little sensor here more one little software patch better here probably will tell you that I'm right now on high alert because I'm in an interview situation. So, can the experience be changed? Friday evening, I've chilled out and all that. Can that experience be be changed? That probably is the next. How that is the next, don't ask me. I'm not uh I would not know it. All I'm saying is that technology keeps on giving you superpowers, boss, right? And then you find a ways to to do it. Again, going back to my cricket uh match experience example, earlier the ads were at a pretty defined space. Somebody figured out you know over beach may ad somebody figured out electronic billboard P and then somebody figured out if it is electronic billboard then people seeing in Singapore versus seeing in India will see different electronic billboards even that is also right all of this is evolution in my your time right >> so I think technology gives us that superpower right for us to go beyond what is visible today >> interesting since you threw in uh those terms and we I've been talking to different marketers about so we uh about quantum marketing nano marketing and you talked about neuro neuromarketing I think >> quantum is a big thing by the way >> quantum yes uh emotive marketing right how >> which is part of this neur neuro huh >> neuro how how something can read your emotions and then suggest contextualize basis that Right. >> So, so, so interesting conversations lie ahead. I think this was also a very interesting insightful conversation. I think what I do take back is that while we are evolving the basic tenants still lie in that foundation. So, >> so I I I I I think the the fear that marketers have that they'll become redundant. I don't think I think it's just a need to evolve as a marketer and I think people who who who will evolve will sustain as is the Darin's law. So Darin's law of nature applies here as well. >> Very good point. Very good point. Darwin's law absolutely plays. Good point B. That is my takeaway. >> Interesting. Thank you so much for uh giving your time and we'd love to have another interaction with you maybe uh a couple of months down the line >> and hopefully the conversation again >> fortunately or unfortunately the conversation might be completely different >> it better be those are because that is the interesting things we live for bridge >> interesting thank you so much for uh giving us your time and thank you viewers who are watching >> thanks for being patient with me thanks for having me here bridge it was great conversation thank