Why should you take a “business-problem-first" approach instead of a “data-first” approach? How do you break down silos within large organizations? In this episode, host Ian Cook sits down with Deepak Jose, a data and analytics industry leader who has spent time at major brands like Coca-Cola, Asurion, and Mars. Deepak shares his thoughts on successful operating models and the responsible use of AI.
Why should you take a “business-problem-first" approach instead of a “data-first” approach? How do you break down silos within large organizations? In this episode, host Ian Cook sits down with Deepak Jose, a data and analytics industry leader who has spent time at major brands like Coca-Cola, Asurion, and Mars. Deepak is a thought leader in data analytics and the CPG space. He shares his thoughts on successful operating models and the responsible use of AI.
---------
Key Quotes:
“Data and analytics should be in the DNA of everybody in the organization.”
“We decided we have to start with the business-problem-first kind of mindset. So what is a business-problem-first mindset? You need to have a very clear understanding of what you are going to solve, how it is going to get embedded, and how this is going to make a decision better for your organization.”
---------
Show Timestamps:
(01:39) What does One Demand mean for Deepak?
(02:30) Creating a single source of truth
(06:21) How do you break down silos within a large company?
(08:30) How do you get each team what they need to tackle their problems?
(10:30) What is better than a data-first strategy
(13:35) The DVF framework
(16:12) Does Mars want their business teams to have data skills, and vice versa?
(17:54) How are they thinking about generative AI at Mars?
(23:59) Responsible usage of AI
(26:03) A cost-centered mindset to a profit-centered mindset
(30:15) Should data teams be separated from IT?
(32:45) Importance of data-backed decisions when navigating changing data trends
(37:15) Closing comments from Deepak
--------
Sponsor:
This podcast is proudly sponsored by Seek, the leader in cloud-based creation and delivery of industry-focused insights.
--------
Links:
Connect with Deepak Jose on LinkedIn
Connect with Host, Ian Cook on LinkedIn
Follow Seek on LinkedIn
Ian Cook: [00:00:00] Some days, it feels like we're drowning in data, but data isn't knowledge. Data without context, or the expertise to understand it, is just something taking up space in your warehouse. That's where being insights driven comes in. Insights driven organizations focus on generating actionable insights, rather than just collecting and munging data.
Insights drive better decisions. Welcome to the Insights Factory.
Welcome to the Insights Factory. My name is Ian Cook. I'm your host and the CTO of Seek. Today on the podcast, we are thrilled to have Deepak Jose. He is an internationally known leader in data and analytics, a thought leader across the industry for CPG and retail. His current position is Global Head of One Demand Data and Analytics at Mars.
We are thrilled to have you, Deepak. Thank you so much.
Deepak Jose: Thank you very much, and I think I'm [00:01:00] really thrilled to be here. Um, and, uh, as part of my journey, uh, at Mars Wrigley, I am very thrilled to be, uh, uh, be part of the One Demand Data and Analytics Solutions team. Um, my team focuses on building data driven decision making capability, data products to drive value for the organization.
So, very excited to be here, and it's always thrilling to represent the team on behalf of a fantastic group of people.
Ian Cook: Fantastic. Well, we're thrilled to have you. I do want to start right there, though, with your job title. It's really interesting. One Demand Data and Analytics, and one of the things that you're known for is working on data as well as insights.
What does One Demand mean for Mars and for
Deepak Jose: you? So, One Demand is a broader transformation program within Mars is by giving you a metaphor.
[00:02:00] You might have heard this quote from Sam Walton, the goal of a company is to give the customer service. Not just the best, but legendary. Now, what he told several decades back, it still holds true. And for March Wrigley as an organization also, our biggest aim is to give the best integrated brand experience.
Now, uh, this is where my metaphor comes in. You might have heard of, uh, four blind people explaining an elephant. Uh, somebody touching the trunk, somebody, uh, touching the leg, calling it something else, somebody touching the tail and calling it something else. It goes on and on. For us as a CPG player, or a brand company, the elephant in the room is the consumer.
And sometimes in large organizations, the sales explains the elephant a certain way, the marketing explains the elephant in a different way, the supply chain, the manufacturing. [00:03:00] So how can we have one single source of truth when we explain the consumer? And that's where we as an organization decided that We need to stop creating silos between the sales function, marketing function, and every other function, and we need to build a solution which is going to be built on single source of truth.
That's what we call as One Demand Data and Analytics, and that is the way we have built the team.
Ian Cook: So it's your group, so it's your team building that single source of truth. You become that underlying point of reference to understand what that elephant looks like. The customer, in this case, and the, you know, it's not a snake because you've got the tail.
It's not a tree trunk because you've got the leg. So you're working on this.
Deepak Jose: And that one single source of truth, what we are calling is the Connected Data Foundation. Now in large CPG companies, Uh, uh, probably for several decades, we had two sources of truth, right? Like one, so two sources of [00:04:00] data. One is the internal data, finances or like supply chain, a lot of internal source of, uh, data.
The second biggest source of data was third party data, which had Nielsen, IRI, or large... Third party data providers used to provide. Now, in the last decade or so, there are new data sources which have started emerging. Let's start with the zero party data, the first party data, and the second party data.
Now, the zero party data is essentially... The data which the consumers are willing to give by proactively to brands and retailers. Is that data they
Ian Cook: are giving, like you said, um, through their activities in a store or is this something they're offering when they buy things to say like, here's my information.
Yeah,
Deepak Jose: so I think, uh, there are a lot of consumers who, uh, are willing to give this data when we have campaign programs or, like, interacting consumers when they come online through some of our websites or some of the [00:05:00] activation, et cetera. What consumers proactively give. Now, another aspect is the first party data, which we get the data through, say for example, M& Ms.
com, the easiest example that I can give. The second party data is the first party data primarily that we get from a retailer or through a loyalty program, which these programs or retailers are willing to share with a brand like Mars, which can be used. So,
Ian Cook: so it's first party. When it's coming from something you own, like M& Ms.
com, it's collected from the consumer. Second party is it's collected, but it's not collected by something you own. It's a retailer through which you are selling M& Ms, say.
Deepak Jose: Exactly, exactly. So that's the best way to articulate this. Now, for us, the single source of Truth is essentially an integration of zero party plus first party plus second party plus third party with the internal data source.
[00:06:00] Now, is it going to stay like this probably after five years? Absolutely, there might be some other source of truth which is going to help us build that single source of truth. And for us, building that connected data foundation is the foundation of building the One Demand story for Mars.
Ian Cook: That to me seems like you have to have a whole lot of people in the room, and you've mentioned breaking down silos.
How do you get those silos working together in one place when you talk about this breakdown of silos? What is going on to actually cross all of those business
Deepak Jose: lines? Yeah, so I think if you think about building an integrated brand experience, see, I think the domain expertise is needed, but this kind of questions that we are being asked would include multiple cross functional things.
I'll give you an example. So, let's say you are a general manager for a market, and let's say if you are getting, most of the GMs are asking this, If I have one [00:07:00] additional dollar to invest, where should I put it? Would I put it in trade? Would I put it in media? Would I put it in shopper marketing? Would I put it in consumer promotion?
Would I put it in digital commerce? So there are a lot of, I'm just giving you one example and There are capabilities which break the silos. Their biggest advantage is they are able to answer these questions. Another popular example, right, like when you go to one of your retailer websites or one of your pre op play websites.
So, let's say, let's, we go to Amazon. com. So, we have to win in search to win at Amazon. So, we go through a lot of algorithm optimization. We might be buying keywords, or we might be doing, uh, spend on the platform to win in search. But, let's say, think about the fact that if the supply chain data or the out of stock data is not there, we would be Spending dollars to win in search, and let's say if that is out of stock, we would have done the, [00:08:00] uh, investment, and the investment, uh, goes to waste.
So how can we have that single source of truth by breaking down the silo? I mean, these are some of the examples. And it is imperative, especially in this omni channel world, for multiple functions to come together. And solve the business problems, which is most impactful for the business. So that's the way we think about it at Mars.
Ian Cook: That's really interesting. One of the things I think of that might introduce, that are a set of problems to tackle or opportunities, if you're a more optimistic speaker than me, like every problem is not a challenge, it's a challenge or an opportunity. Each one of those groups, like you mentioned, supply chain, marketing, whatever it's going to be, is going to need their own set of insights from that connected data foundation.
How do you make sure that each one of those teams has what they need? Because it's not going to be the same, I don't think, that each, the rest of the half, each of the rest of them need. But it is coming from the same [00:09:00] source of truth. What is that process to make sure, say, supply chain has the kinds of insights that they need to say, all right, great, this is how I can make progress.
We go to marketing, this is what our customers look like,
Deepak Jose: etc. Ian, this is a wonderful question. So at Mars, uh, So in many organizations, uh, including me in the past, we have always told that we have a, we used to tell them that we have a data for strategy, right? And these days we are, uh, changing our, uh, our operating model from a data for strategy to what we call as a business problem for strategy.
Now, if you are 15 years back, we used to spend millions of dollars in building a data warehouse. Over 10 years, 5 to 10 years back, we started building data lakes. These days, we are building lake houses. In the future, we might be building, uh, I mean, data mesh, right? So, it's always a data foundation. And sitting on top of that, I mean, leveraging the data, we would generate some knowledge and insights.
[00:10:00] And, uh, leveraging this insight, we would take some actions, which would create some value for the company. Now, My provocation, and this is what we have done at Mars, is we wanted to invert the pyramid. We decided we have to start with the business problem first kind of mindset. So what is a business problem first?
You need to have a very clear understanding of what you are going to solve, and how it is going to get embedded, and how this is going to make a decision better for your organization. That's what we call as the business problem first kind of mindset. Once that is implemented, if you have a business problem first mindset, then you can understand.
What are the actions you need to take to create a value and to generate those value or take those actions? What is the knowledge and insights that you need and for generating those insights, what is the data you need to have? So instead of having a data first strategy, We are operating in the business problem first, data second kind of a strategy.
That is going to be very important. That business [00:11:00] problem first strategy is going to help us prioritize our efforts and we can have a decision back kind of an approach and enable some significant decisions for the company. I think that that is the biggest operating opportunity for us in my opinion. I think
Ian Cook: that's really interesting.
In a lot of places I talk to, there is a push to always talk about being data driven. And I've always wondered if that's the right order in which to go through things. And it sounds like perhaps you're doing things in a slightly different way. Would that be
Deepak Jose: correct? No, I think, uh, so the key theme is like, it is not essentially, the decision making should be always data driven.
But it is not the data... First strategy. It is a business problems first strategy, right? Like if you want to take a data driven decision, I think it should be a business led process rather than a technology first process. Because technologies can come and go. There will be new technologies which will be coming.
New data sources will come and go, right? But the key aspect is, we are in the [00:12:00] business of solving business problems. Uh, we are in the business of creating data products, uh, that people can use and take better decisions. So that is, that is the, uh, I mean, data driven is the goal, but I think business problem first is the operating model.
Ian Cook: So when you work with these other groups, is it the case that they come to you with their business problem? Do you help them think that way? One of the patterns I've often seen from a lot of groups is the data team at any company becomes this sort of ad hoc query machine. You know, they come, they show up, and they're like, I want three years of data about this, and just hand that to me.
That sounds like a lot of kind of a pattern that you have to work on people with, in my, in my. Correct in that? Or have you found that a lot of people really adopt this business? Oh, absolutely. Absolutely.
Deepak Jose: And there are some, I mean, some of the capabilities our organization have built, the business team have built is, it is some of the best in class in the world, right?
[00:13:00] I mean, and it depends on the, and obviously there are areas where we need to make a lot of progress as well. Now, one thing I am personally proud of is some of the work that we have done in the revenue growth management space. See, working with the business domain leaders, we have identified some of the top business questions that we need to solve.
And we have built data analytics capability. To, uh, enable solving these business problems, how to take, uh, how to optimize the promotions, how to do trade architecture, how to do strategic pricing, etc. All of these, depending on whichever domain we work with, at Mars, we use something called a DVF framework, desirability, viability, and feasibility framework.
All right, that's new to me. Yeah, so the DBS framework that we use at Mars and we keep users at the heart of whatever decisions that we do. Uh, like, so what does, what, what is the desirability, feasibility and viability depending on, uh, [00:14:00] understanding the user needs, we'll be able to, uh, create the right kind of tools.
for the right kind of persona. So that's been our experience. Now, at the same time, it is very important to have people who have a very dynamic and a solid understanding of various data sources as well. New, new sources will come through, new technology will come through. Um, so, uh, for me, an analytics practitioner should have a combination of three skill sets.
The math skills, The technology skills and most importantly, the business skills. So, a combination of these three can be very good and it can help us drive some competitive
Ian Cook: advantage. That sounds like a tough person to find. Um, when you say the business skills, do you mean like deep knowledge in CPG? Or maybe something, can they come with, well, I know how to frame some business questions.
How far into the business do you think those, um, analysts
Deepak Jose: need to go? Mostly the, I [00:15:00] have two kinds of, uh, uh, I mean, there are different kinds of associate, but I think broadly two career paths that I've seen is I've seen people who are, who have been in sales or category management, who were really embedded within the business, who have developed their passion in data analytics.
And, uh, they have developed themselves to leaders in data analytics. Now, uh, the, the, I mean, these folks are, uh, I mean, they have gone through the learning of, well, uh, learning of understanding what is the fundamentals of data sciences, data engineering, et cetera, as they have grown through to this role.
But I've also seen, uh, data scientists, the, the new gen data scientists who have been doing their job really well, and they have. Developed or build their passion to work very closely with the business. They've learned the business problem really well and they've built their expertise. Now, one thing, [00:16:00] it is undoubtedly, I can say, it is difficult to find these two talents.
But I think it is all about our ability to give associates a career path so that we give them, we help them to reach a path like this.
Ian Cook: Is there any opportunity or have you had the chance to talk with the other side of the house, for lack of a better term, about going the other direction? Do you often require that your business experts at Mars have some data skills or some technology skills enough to at least have a conversation about the data that they're looking for, understanding the structure of the data?
Or do they generally come to you much more with, look, I know the business problem I'm trying to solve. But, you know, you tell me what data set is out there and what fields you need. I'm, I'm, I'm washing my hands of it.
Deepak Jose: Yeah, it's a com, definitely a combination of both. And I'm, I think some of the, my favorite business product owners have a phenomenal [00:17:00] understanding of what is happening in the data space.
These are the leaders who. Asked us, hey, when the generative AI wave was coming, the large language models were being used. They asked us, hey, maybe you should leverage, here is a manual automation task that we are seeing. Maybe you should build a LLM to solve for it. So there are business leaders who are in depth in business, who also have the Point of view on technology.
I mean, I mean, I have, I've, I've seen that happening. It's, and that is not at all unusual. Now you will also see leaders that, I mean, uh, they have a fantastic understanding of the business, but when it comes to data. We would have to step in and be their partner, uh, while solving the problem. I mean, I've seen, uh, in both spectrum and everybody comes with their own strengths is what
Ian Cook: I would say.
Fantastic. Well, I would certainly lose my podcaster [00:18:00] license being a data oriented podcast if I wasn't going to ask about generative AI. Fortunately, you teed it up for me. Let's start very broad. How are you experimenting with generative AI at Mars? Yeah,
Deepak Jose: so as I start talking about Generative AI, right, like, I mean, uh, see, responsible usage of Generative AI is really, really important for Mars.
Mars is a purpose driven organization. We say the world we want tomorrow starts with how we do business today. So it's really important for us to have a responsibility strategy and how we are using Generative AI or any other technologies. We partner with companies like Microsoft, uh, Responsible AI Institute to make sure that we follow the guidelines.
We have internal benchmarks and standards, so that is something that we are doing. When it comes to Generative AI experimentation, we have built a framework. Um, and I think, uh, uh, and the framework that our team is [00:19:00] using is called Do Things, Do Things Better. And do better things. Now, uh, depending on the, the disruption which each of the capabilities are causing and depending on the value it can generate.
So value and ease of implementation. This is how we have created these three, three broader buckets. So what is, uh, do things? Do things is how we can... Automate a lot of manual tasks, which, which, which might be happening like this can be a code automation or like this can be, uh, a chatbot, et cetera, where a manual repetitive task can be taken out.
Leveraging, uh, the power of lms do things better. Would include areas like how can we, uh, leverage, uh, LMS for a content summarization. Right? Or, uh, uh, creative, new creative, uh, uh, creating new or, uh, content in, in [00:20:00] general. Uh, conversational AI could be another example of do things better. Yeah, so those are some of the examples.
Now the final, the do better things are areas which can be truly transformational. It would take a little bit more time. This would include... Areas like new product innovation, for example, or another example could be, uh, hyper-personalization. These are areas which would take more time. Now as Mars and, and I, I am only part of Mars one demand data and analytics solutions team.
Right. And
Ian Cook: Mars is a huge organization with, it's,
Deepak Jose: it's, so, it's a much bigger conversation. There are a lot of capabilities. And for us, uh, we are doing a lot of several controlled experiments to figure out what are the right proof of concepts so that we can take these models to production at the earliest.
I mean, that's what we are focusing on. Um, one of my biggest fear is, I mean, we can do a lot of experiments. Matter of fact, uh, we can have a [00:21:00] graveyard of POCs after a period of time. How can we scale it is going to be very important. And another important aspect is, like, how can we do it in a cost effective way?
Because, uh, we, uh, working with many of our technology partners, we need to answer the question, should we use the open loop, uh, uh, LLM? Or should we have, I mean, customize an existing LLM? Or should we create an LLM for ourselves? The answer could vary and with the technology accelerating really fast, I think, um, maybe what I think today might not be the answer in one year, right?
Like, so, yeah, I think a lot of companies
Ian Cook: are going to go through. Whether they should build, in your case, Mars GPT, or just say, you know what? There's a company with several hundred million dollars working very hard to keep ahead of the curve. Maybe we keep just relying on their extensions and fine tune [00:22:00] on our data, which is an approach I know a lot of places are taking, but it's a very hard balance to do.
Of course, we're not getting into any business secrets here, but when you say you do all these experiments, do you have, is One Demand part of a group that is doing these experiments?
Deepak Jose: Absolutely, absolutely. We have a data sciences team and a AI team which is focusing on this on a day in day out basis. Now, I think as a data analytics practitioner, one thing which gives me a lot of happiness and energy is the pace in which the technology is changing.
Okay, so this is, this has been quite exciting for me because I feel like you can do a master's in a period of three months because the new capabilities coming out, the new technology coming out, and I think it's been quite exciting. There are a lot of changes which is happening. I truly believe that generative AI is.
Really transformational. Now, [00:23:00] this has been called everything transformational, but I think, uh, uh, as the building of AI revolution has begun, it is all up to us to make the most of it. What,
Ian Cook: not only in the LLM space, but as a practitioner and as somebody who's looking forward, what are you seeing that you're really, really interested in?
Deepak Jose: I think, uh, see, one thing which I'm personally quite excited about is, uh, Uh, I mean, under the guidance from White House, the Microsoft team is collaborating with Amazon and Google and Anthropic and OpenAI to, and putting a council to make sure that there is a responsible AI usage, and I think that has been truly, uh, I am personally excited to see that partnership at the technology sector level, and I fundamentally believe that responsible usage of AI is going to be detrimental [00:24:00] for the future.
What is responsible? Yeah, yeah, the responsible usage of AI.
Ian Cook: So, what does responsibility mean for you in using AI at Mars?
Deepak Jose: Yeah, I think now, uh, I'll give you my broader theme, and I think this, uh, there are several principles which Mars is... Uh, thinking about and as part of our Responsible AI policy, and we are, uh, hoping that, uh, uh, we will be publishing some of it for the whole, uh, industry, uh, in the, in the future as well.
But some, I can give you a quick summary of some of the key things, uh, you would see from our, uh, the latest press release on Responsible AI. The one is on transparency, right? Like when it is important to make sure that. Uh, the models and the transparency in the models. It's, it's, it's something really critical.
Uh, we need to make sure that we take out unconscious bias, biases in the models. That is another important aspect. [00:25:00] Um, the explainability. Uh, is another, uh, important thing that we are looking at. Uh, privacy, security, uh, and collection, when it comes to the collection of data. Now, Mars, uh, I mentioned that like Mars is a responsibility first company.
Mars is an organization as part of a responsible marketing code. We don't collect data from children under 14, right? And that is the foundation of how we want to do business. So I'm really passionate about our responsible AI team and the responsible marketing team. And it's always, it gives me a lot of energy to work for a company which wants to do the right thing for the society.
That's
Ian Cook: terrific. I can talk AI all day. It's the field I came up in and it's fantastic. But I do want to get back to something that I picked up on in some of your writings and discussions that I thought was really interesting. I'd like to bring you back to a distinction [00:26:00] you made that I think has real importance that we would like to dig into.
You said that data teams should move from having a cost centered mindset to a profit centered mindset. Can you elaborate on
Deepak Jose: this? Absolutely, absolutely. Now, I mean, this is closely linked to the conversation that we had earlier on. The business problem first or a data versus a data first kind of mindset.
Say, let's say I hypothetically, I'm a data first mindset guy, right? I'm building a huge data lake, which, and I'm building APIs to ingest, I mean, petabytes and petabytes of data, right? Now, let's say I have this conversation with the leadership of the company, And when the optimization of my resources would always mean, hey, how can I do the same job at a lower cost?
I mean, that is one way of looking at optimization, right? [00:27:00] Now, if you are moving from a business problem first kind of a mindset while building data products, instead of a data ingestion strategy or like data first strategy, you will be able to ask the question, hey, if you invest this dollars, I might be able to give...
3x return, or a 5x return, that is the bigger conversation that we need to have. Now, the difference of doing this mindset is, and the advantage of doing this is that when we do it right, we'll be able to add a lot of value for the company. Now, instead of cutting costs, When we are driving revenues, improving profitability, I think that's the right conversation that we need to have.
Uh, that's my, uh, that's my thought process and I, I hope, uh, the industry starts thinking this
Ian Cook: way. So that's an interesting question is getting people to start thinking that way. Tell me about experiences you've had in trying to get people to think that way that have gone well, or maybe not as well. What are the common [00:28:00] bumps in the road that you
Deepak Jose: often hear?
I think, um, I'm, I'm seeing a lot of industry leaders are leading this trend. Like, I mean, the value first mindset is not something patented to Mars or patented to me. I mean, I've heard. Several industry leaders talking about that business problem first mindset, so I think this is not new. Now, see, I think if you think about the life of a data scientist or the position of a data scientist, Probably, uh, almost a decade back, right?
So data scientists or like people who work with data, they were always considered as somebody who sits in the back office. It was a back office function. It was, uh, it was too far from the decision making process, right? Like when, uh, that's the way it was developed. Now, if you, if you think about generative AI, like the day of today, Think about the amount of conversations C-level executors are having about generative ai.
Uh, [00:29:00] we have, uh, one of our senior data scientist leader in the company who has a lot of experience in, uh, N L P and M L Q, and the, the number of conversations where he gets asked questions about generative AI is like, skyrocketed, right? Like when, uh, uh, I think data and AI part, uh, data and AI teams. Have a seat at the table today, and I think the way technology is getting involved, I think there is a bigger role data and AI leaders can play.
Now, uh, and we have to be humble in the process. We need to treat this with a lot of empathy. There should be a very good understanding of organizational culture. We need to have a very good understanding of the change management that we need to go through. Because building a tool is relatively easier, but I think we need to spend for every dollar that we, uh, we spend in building a tool, we need to spend an additional dollar when it comes to maintaining not only, [00:30:00] uh, I mean, training and upskilling, change management, uh, how can we, uh, Broadly, elevate the data driven culture in the company, uh, with leadership support, with storytelling.
So I think that is going to have an exponential impact. To
Ian Cook: achieve that, do you see that the data organization needs to be different or separate from, say, the IT organization? And the IT functioning? Part of my job, I've had at different companies, it very much is that cost driven centers. Like it's, I, my optimization is for the slow, lowest AWS bill I can get, how much can I do?
Versus the more of a business data driven side saying, what can I do in terms to, in terms of, um, multiplying the work marketing's doing, multiplying the work that the supply chain is doing. Both of those seem relatively important. Perhaps they need to be all together in one. Maybe that's the way
Deepak Jose: you've worked on it.
I think I [00:31:00] have seen both of the models working and both of the models working well in different organizations. So I think the org structure is not the key thing, but the effectiveness of the operating model is the more important thing. So, for example, during my stint at Coca Cola, I mean, a lot of the data analytics function were sitting within the business unit, reporting to the business.
My current role. Um, I report to, uh, my leader who reports into the chief growth officer of the company, right? Like I said, it's a very business focused function. Now, uh, we have seen a lot of aspects of it is working really well. Now, I've also seen, uh, in my previous role at Mars, we were reporting into the chief information officer function and it was also working.
So I think it has everything to do with the operating model. Not specifically focused on, uh, the organization structure. You can have a really successful strategy getting implemented, sitting within the IT function [00:32:00] or within the business function. I think the execution of, and every organization has different operating models as well, right?
Like when, so it's, it's... Very dependent. Now, I did, uh, my, uh, uh, brief stint in consulting as well. Now, the, the last thing that you should do is, like, you take a boilerplate template, and go to one organization and tell that, hey, this worked in a different organization and you should adopt it. I wouldn't recommend doing that.
I
Ian Cook: think, uh, That seems to me to be what a lot of this discussion about data mesh is as well. You brought this up earlier. It's like, you know, fine, we have a warehouse. Now we have a lake house. So that's terrific. But, um, a ton of these things are passing fads and they seem to want every company to be like, the answer to your problem, data mesh.
Now do that.
Deepak Jose: See, I think, uh, so this is, this is, this is where my passion is on a decision back approach or a business problem first approach, right? Like when the same organization ha can have a common data layer [00:33:00] where a semantic layer can work or a data mesh can work, it can be a federated. Operating model, or it can be a centralized operating model.
Well, it depends on the business questions. And how can we enable the decision maker? So I think that should be the focus. Because, see, I mean, if you are in the data and analytics space, sometimes we'll get centralized, sometimes it will be decentralized, sometimes it will be a hybrid model. But your ability to be Agile and your ability to adapt is going to be, uh, very important, uh, in this process.
So I think, uh, I mean the, the moral of the story is that I think if you're willing to adapt, if you're going to be agile, then you'll be successful in this journey. That's
Ian Cook: terrific. Deepak, you're always in demand. You're an author, you're a speaker, you are obviously heavily engaged across Mars in what must be one of three day jobs.
What are you working on now? What are you, what can we look forward to from you in the near future? [00:34:00]
Deepak Jose: So, I think before I talk about what I do now, so I think you might have heard of this concept, Japanese concept called Ikigai. And for the listeners, if you have not heard, you should definitely check about how can you build a happy life, right?
Like, I mean, a purposeful life, uh, with your, uh, that, that's the theme of Ikigai. So Ikigai essentially tells that, uh, you should be doing what you are good at, uh, what you are passionate about, what the world wants. And what the world is willing to pay for. And for me, uh, my current role at Mars is in the intersection of a lot of these things.
Uh, and I think it gives me a lot of energy to do this, uh, because of that. So, and one of my personal passion is to help students. Who might not be, uh, who wants to become a data professional. I, I want to inspire them [00:35:00] to pursue data analytics as a, as a professional. And then I get a lot of energy in that.
So I spend a lot of time talking to students, uh, to influence them. I think that if I can even address one student who is confused between three different career paths and if I am able to influence them to follow their passion in data analytics and AI, I think I would be really happy. So I spend a lot of time doing that.
A lot of things that I'm working on now, uh, uh, within the organization is on and within the external team is also on. How can we drive the adoption of, uh, data analytics and AI in a better way? Uh, I still see that in a lot of large organizations. While we have a lot of success stories, uh, there, there is a significant opportunity for us to simplify what we have done as data analytics practitioners.
So that data analytics is not the job of [00:36:00] analytics team. Data analytics should be built as part of the DNA of everybody in the organization. So a lot of exciting conversations. In those regard, which gives me a lot of energy.
Ian Cook: Fantastic. Well, Deepak, before I let you go, I want to say I've had a fantastic time talking with you, everybody who's listened to this is going to want to hear more from you.
Where can they find you or where can they find more of your writings and the presentations you've given?
Deepak Jose: Yeah, I think, uh, I would have liked. to say, I think you should follow me on threads, but I am not active on threads. I'm, I'm, I'm quite, uh, you can reach out to me on LinkedIn. You will find me as Deepak Jose.
I think that's the best way to keep in touch. And I, uh, I make it a point that if there are, there is something significant happening in my career, or if I see something, some. Cool work I try to share, and that's the one way I want to give back to the community. Uh, so that's the best way to keep in touch.
I look forward to connecting with a lot of your [00:37:00] audience, and I would love to have this intellectual conversation
Ian Cook: going forward. Well, I know I have thoroughly enjoyed it, and everybody else is going to love hearing from you as well. So best of luck with everything you're working on. Hopefully we can have you back again, and maybe the next time Google releases a chat GPT killer, you can come on, we can talk about how that's going to transform everything.
Deepak Jose: I think I'm, I'm looking forward to it. And my closing comment to your audience, right? Like I grew up in India, so a lot of my leadership principles are influenced by mother Teresa. So she once told this thing, not all of us can do great things. But all of us can do small things with great love. So my call to action to all of you is find the small things you are really passionate about and do it with a lot of love.
That is going to change the world. And that's how you can. Hit the ball out of the park. So at Mars, we say the world we want tomorrow starts with how we do business today and your tomorrow starts today. So have [00:38:00] a, a nice, uh, time and find your things, which you are passionate about.
Ian Cook: Beautiful. Thank you so much Deepak.
Have a great day.
Deepak Jose: Thank you again. Thank you for giving me this opportunity.
Ian Cook: This podcast is proudly sponsored by SEEK, the leader in cloud based creation and delivery of industry focused insights. Thank you for listening. If you liked it, please feel free to rate and subscribe wherever you get your podcasts.