Data Analytics Chat

Why Have Firms Not Focused on Data Readiness

• Ben Parker

Share episode

Send us a text

In this episode of Data Analytics Chat, host Ben Parker interviews Nipa Basu, a seasoned data analytics executive. Nipa shares her captivating career journey from pursuing a PhD in economics in India to becoming a Chief Data Officer in the United States. They discuss pivotal moments that shaped her career, the evolution of data science, and what distinguishes good leaders from great ones in the tech industry. They also delve into why many firms struggle with data readiness and the importance of balancing technical expertise with soft skills. 

In this episode, you will learn:

  • Nipa Basu’s career journey from technical contributor to C-suite executive in data analytics
  • The difference between good and great leadership in data science
  • The importance of storytelling and communication skills for data professionals
  • How to build and lead large, high-performing data teams
  • The evolving landscape of data science and the need for continuous learning
  • Why many companies struggle with data readiness before launching AI initiatives
  • Common misconceptions about data readiness and the role of the C-suite
  • Practical steps for organisations to become truly data-ready
  • The impact of organisational culture, skills gaps, and leadership on data success
  • Real-world examples of career setbacks and adaptability in the data industry


00:00 Introduction and Early Career

01:37 Welcome to Data Analytics Chat

02:22 Nipa Basu's Career Journey

03:23 Leadership in Data Science

05:20 Navigating to the C-Suite

00:00 Transitioning Between Roles

21:30 Building and Leading Teams

27:07 Challenges in Data Readiness

41:48 Practical Steps for Data Readiness

45:00 Conclusion and Final Thoughts

Thank you for listening!

Nipa:

So I came from India to United States to do a PhD in economics, and I was doing that in the State University of New York at Alban. And like many other grad students, I was also working for the state government, albanese, the state capital, so a lot of people do that. I was working as an economist and as I finished my PhD, my first post PhD job was for a defense laboratory. This was a bit unusual, and I'm of course talking about ages ago working for the government for economists is quite common, but not so much in a defense lab. It was the Sandia Defense Laboratory in Albuquerque, New Mexico, and there was a visionary nuclear physicist who started this program of creating a micro simulation model of the US economy. Sandia Lab had the world's largest, massively parallel computer, and we were putting agents on the parallel nodes of that computer, and trying to simulate their behavior. The project was just starting, so I joined at the very beginning. What I'm describing today, people will recognize it as AI or data science

Ben parker:

Welcome back to Data Analytics Chat, the podcast where we discuss the world of data, ai and the careers shaping it. I'm your host, Ben Parker, bringing you real stories, expert insights, and practical advice to help you thrive in the data industry. Today I'm excited to welcome NEPA Basu Data Analytics Ticket Executive. In this episode, we'll explore her exciting career journey. Pivotal moments that shaped her career path. And discuss the data topic on why firms are not focused on data readiness. Nepa, welcome to the podcast.

Nipa:

Thank you, Ben.

Ben parker:

Cool. I look forward to today's conversation.

Nipa:

Me too.

Ben parker:

So do you wanna start off with introducing yourself and sharing your career journey?

Nipa:

Of course. I'm Nipa Basu. My career has been devoted to data analytics, and I think it has been pretty straightforward. I'm doing data analytics all my life, starting first as a hands-on technical contributor. Continuing first to leading small teams. Ultimately, it led to being part of the C-suite and also I had roles of leading practices, but it has always been about drawing insight from data using analytics, starting with the New York State government, then a defense lab. Att, then Dun and Bradstreet, and lastly.

Ben parker:

Brilliant. And so what differentiate a good leader from a great one.

Nipa:

Not sure, leadership can take so many paths. Especially this path of data science that I took there are folks who remain true experts. They're still innovators doing hands-on work, still building new things. That's a great path and I have a lot of respect for them. That wasn't my path. My path was to never lose touch with the core of data science and then moving on to leading teams, leading large teams, and then getting the results working with my team. So I think one differentiator is if you depend so much on your team that you really have to know how to influence them and how to inspire them. At the same time, you still have to have working knowledge of something like this, which is evolving and almost changing every day. I think that's one. Then the other side of it. You have to be, this is a highly technical field, but you have to be able to tell the stories associated with this highly technical field to others who are not so technical, especially to the C-suite. If you don't get buy-in from the C-suite, then you will really not be able to do much. So I would focus on those two.

Ben parker:

Okay. Interesting. And I guess obviously you know, you've been successful with your career to make it to C level and there'd be listeners who got at leadership level already or lower down the sort of their careers. How did you like navigate your journey to become a CDO? Did you always want to have that goal in mind or did it just gradually progress?

Nipa:

Yeah. Yeah, so it's, I had a C-suite role. I won, say I had that in mind when I started my career. As I said, in the beginning I was highly technical and I was primarily focusing on the technical role. Maybe, if you don't mind, I can share the story of how I get involved in this field, because I think that's. that okay if we go there?

Ben parker:

Yeah, please do.

Nipa:

Sure. So I came from India to United States to do a PhD in economics, and I was doing that in the State University of New York at Alban. And like many other grad students, I was also working for the state government, albanese, the state capital, so a lot of people do that. I was working as an economist and as I finished my PhD, my first post PhD job was for a defense laboratory. This was a bit unusual, and I'm of course talking about ages ago working for the government for economists is quite common, but not so much in a defense lab. It was the Sandia Defense Laboratory in Albuquerque, New Mexico, and there was a visionary nuclear physicist who started this program of creating a micro simulation model of the US economy. Sandia Lab had the world's largest, massively parallel computer, and we were putting agents on the parallel nodes of that computer, and trying to simulate their behavior. The project was just starting, so I joined at the very beginning. What I'm describing today, people will recognize it as AI or data science. We were not calling it that at that time. We called it I think, computer science, and I was an economist lucky enough to get involved in this very interesting project. I could stay there only for two years and for family reasons, I had to move to the East coast to New Jersey. I started doing more traditional analytics. I, I joined at t at that time at t had very rich data. Then I spent a lifetime in Dun and Bradstreet that also had the world's largest small business data. And initially all my interest was really to get my hands dirty and what I could find in the data. was the goal. I was not looking at corporate hierarchy. I was not looking to get promoted. In fact, in the beginning, I resisted leading teams a little bit fearing that I would lose my hands-on skills. But then it happened over time, first I had a very small team, then a slightly larger team. And then I got lucky because this is the time when in parallel this field of data science was growing and whatever I was learning in Dun in Bradstreet or at t, I could find links to what the computer scientists were doing in that laboratory. So I think initially I differentiated myself as a highly technical person. So my teams wanted to work for me. They also wanted to work for a highly technical person. And then gradually I saw that perhaps my translating skills and storytelling skills are better than others. So gradually I became. Zo with the businesses and started contributing to business decisions. And I think I was, I became part of the business, make business decision making teams so easily because the company was so focused on data analytics. Data analytics was their product. So that's how it was. Gradual move and ultimately I think my team leading skills and storytelling skills played a more important role in becoming part of the C-suite. And, this role of chief analytics officer, that itself was new. There were a handful of CAS at the time. So yeah, it was gradual.

Ben parker:

Okay. Interesting. And then obviously I guess a lot of people do have that, like you said, the fear of losing your technical expertise. And I think it's quite common in an industry and obviously at that level you are technically, you really experienced and that's you, your expertise. How did you then start to unfold your, the softer skills, which is obviously important for the leadership aspects.

Nipa:

Yeah. Yeah I think it, it happened naturally and I think for the longest time I used to be the most technically savvy person. In a group of people who are all actually quite good at soft skills and in communications mostly, a room full of MBAs and me. On the other hand, I will be the best storyteller in a room full of highly technical people where everybody's really technically good. But my age was that I could tell that story in plain English. The companies that I worked for also helped me. There are formal trainings and folks who are listening the junior folks who are still trying to improve these skills just the same way. Nobody learns, learn coding just by magic. You have to learn it. Similarly, good communication skills can also be learned and practiced. It is important unless you want to be, the mad scientist sitting in a room and inventing something great and not need to interact. Those are becoming more and more rare.

Ben parker:

Yeah, I agree. And so how did you learn? Was it, do you get mentoring? Was it out of hours, extra study or classes? How did you go around doing this?

Nipa:

All of it a little bit. There were different training programs and most of this happened in Dun and Bradstreet, where I spent 20 years of my career. At different levels. The company invested in it, on my own also a little bit. Yeah, it was a combination of of a lot of small things and I liked to do it. A lot of people don't like to go. To a conference to present. I enjoyed doing that and these things I discovered, I did not know it in the beginning that I will enjoy it. So yeah, it was iterative experimentations

Ben parker:

Yeah. No, I like that.

Nipa:

that. Yeah.

Ben parker:

But obviously you've got, you have to put yourself out there, don't you? Because if there's so many things in life that you might, that you don't do, that you might actually love, it's, you just need to, troll and error, isn't it?

Nipa:

Absolutely.

Ben parker:

So then looking back, were there key moments that you believe shaped your career?

Nipa:

So I don't think those were discrete moments, but continuously what was happening in this field shaped my career. When I chose this field, let's say I was, maybe I was a 25-year-old professional. I did not know where data science was going to be At that time, dealing with data was a nerdy thing. You sit at the back office and you pour through, through numbers. As I was focusing on my career, everything about this field changed, right? It's so very early days. I can tell you another story, so I used to go to economics conferences, right? As a graduate student. And those conferences will be dominated by. Gray haired professors they were always on, on stage. The first time when I started going to computer science or data science conferences, I saw that it was completely different. I was no longer a grad student, but the grad students were dominating the whole thing. Most of the times it was them presenting. And I slowly understood, I think somebody explained it to me that this field is so new and it's evolving so fast. not necessarily true that professors know more than the students. They're also learning and there is not a huge advantage of accumulative knowledge. I think somehow it stuck with me and I realized if I want to stay in this field, accumulation of knowledge is less important. What's more important is being open to the idea that new things will be happening every day, and you will be learning new things every day. That might make your earlier learnings not as relevant. And I continued with that team. It stayed with me as I started leading teams and had very large teams, and I always knew that the junior, most folks are really the most technically knowledgeable folks. So I think there are some differences in this particular career associated with data science, data technology, data engineering, where the things happening in the world shaped our careers sometimes unknowingly.

Ben parker:

And you hear that quite a bit now, the, I think it's the. Especially when you go up to senior level, you obviously you can't know everything'cause there's so much going on. So you need to be able to rely on the team to get your knowledge from. So then obviously they're the technical experts and you're trusting the people you build the team with are gonna be able to provide you the right knowledge.

Nipa:

Absolutely, and they have to trust you, to do their best.

Ben parker:

Yeah, I think it's just important, isn't it, to work as a team have that good bond between you.'Cause everyone's, you're on the same journey, aren't you? You want to achieve whatever your objective is. If you've got a really close knit bond, I think it's will propel your successes.

Nipa:

Absolutely.

Ben parker:

So then have you ever faced a major career setback?

Nipa:

So many I can talk about maybe the most impactful one or the most important one. This was done in Brad Street. This is where I have spent a lifetime. And the team there ultimately was about a hundred people. Analytics team. It's a data analytics company. It had a huge data team. I was leading the team that uses that data to build solutions, and most of those hundred people were actually hired by me and I built the team. But in 2019. The company changed direction completely. A private equity firm came on board. They actually acquired the company, so it stopped being a publicly traded company and was owned by the PE firm and everybody with a c in their title were gone, including the CEO, the CIO, and I lost my job at the same. At the time, it was a really devastating experience. I think I would have stayed there and retired from there if it didn't happen. Now I wish I could say that from that I learned the trick of making sure it never happens in my career again. not true. I did not learn that. When there are turmoils at the top, this could happen again and again. But what I learned within four months, I was leading a practice in GHD. What I actually learned is the change that it brought on was actually quite good. Though I was doing increasingly more and responsibilities were coming to me, but I was still in the same company for 20 years working in the same field though with increased responsibilities. From there, I went to a completely different type of company. GHD. Not too many people may know about it because it's an Australia based company and it's an a EC company, architecture, engineering, construction, completely different from the Dun and Bradstreet world. But what I learned there is the importance of keeping up with your core strength. In your core strength could be about data science. It could be about leading people so that your skillset are transferable and you can fit into a different world, a different company so easily.

Ben parker:

I love it how you've used your adaptability into a new role. I think it's so key, isn't it? Like you just, like you said, you mentioned stick with your strengths.'cause Yeah, that's where you, that's where you're strong at. That's where you need to utilize. And obviously there will be gaps in anyone's expertise. And then that's where you need to lean on other people, isn't it?

Nipa:

Absolutely. And, we are in a very different time. We don't know how the working world is going to change with the influence of ai. Those who are in the data world are working with it closely, and whether you are in the data world or not, different kinds of changes are coming. So it. It'll be very hard to stick to the same things that we are doing today to continue with that, with or without a job loss. So that's why it's important to keep the skills relevant and and up to date.

Ben parker:

Yeah, I agree. It's ever changing. And also I just wanted to go go back on, you mentioned you built the team of a hundred people and that's obviously Yeah. Remarkable. What, how did you. Foster that culture.'cause this obvious challenging job, isn't it? Building a team and I guess building a team and management is completely, or leadership, whatever you wanna call it, is completely different skillset. So how did you go about building a sort of a large team? I.

Nipa:

Yeah. Yeah. In Downing Brad Street, I had a hundred people. Team. And then in, in GHDI actually had a team of about 200, and I've talk about both because the two teams were quite different. Dun and Bradstreet, I built the team almost from scratch. There were probably five or six people who were parallel to me when I started leading analytics and they became part of my team. But then I started hiring and we hired junior folks, fresh PhDs, primarily because of their technical skills and the industries were different. In Dun and Brad Street, it was an established data analytics company. The clients were mostly, financial services, banks insurance companies retail telecom clients were also data natives. They were in that data world and our role was to prove to them they will continue to do business with us. If we can prove that if we can prove that our solutions are better than our company. So there was a lot of focus on building farther the technical acumen the quality of data science and the quality of data. This company's product was data. So it was easy to get investment in those. So the teams consisted of young ambitious folks who are really interested in data and technically efficient. And then I had team leads, of course I did not have a, all of them as direct reports. So the leaders were senior data scientists who also had the. Skills. There was almost nobody in the team who was not technically very good. That was the necessary condition, but they had the soft skills and I was really running the team relying on them. In GHT, the experience was slightly different. It was an even larger team. It was a global team, long before the pandemic I was actually working from home with half of my team in Australia and New Zealand and also in different parts of the United States. And I mostly inherited the large team, so I did not have that, that I hired them, mentored them. Many of them were quite, quite seniors and for many of them, I think it was it was a mutual respect with my directs because I probably had more data analytics credentials than any of them. That's why I was hired. But they knew the industry, a completely different industry more than me. I did hire quite a few data scientists there, but the rest of the data professionals were working in the field for a long time. So you adapt. You learn how to work with a geographically dispersed dispersed team. And you work with people with much stronger domain expertise than you have, and this industry was different also here, we did not have to prove to our clients that our solutions are better than our competition, but we actually had to convince clients to take the data journey to get along the data analytics path. so the skills needed for the folks in the team were also slightly different.

Ben parker:

Brilliant. Two interesting different perspectives there.

Nipa:

Yeah.

Ben parker:

So then what sort of skills do you think have helped you stand out in this industry? Obviously you've mentioned your technical ability but is there anything else you'd like to add upon that?

Nipa:

Yeah, I think I've already talked about it. I think the ability to learn ability to lead large teams, I think people who are far smarter than me, also like to work with me and follow my directions and the storytelling.

Ben parker:

Brilliant. No, and I said I think it's obviously. Apparently, isn't it? You gotta have that, you've gotta have the technical exp expertise and also them soft skills. I think it's crucial. And like I said you've been blessed with these amazing storytelling skills and I think that is quite critical.'cause like I said, if you wanna get stakeholder buying, you need to just be able to translate the technical requirement into a business context, just easily. So it's easily transferable and easily to understand.

Nipa:

Right.

Ben parker:

So let's move on to the data topic then. And we're gonna discuss why are firms not focused on data redness. Wh why do you feel that for many firms underestimate the importance of data redness before launching let say analytics or AI initiatives?

Nipa:

Yeah. Actually lemme divide most of the companies into three categories. And the first two of them do have data readiness. And the last one, which unfortunately is the vast majority of companies, does not for the three categories. I will say the first category is. companies whose product is data or analytics or ai, like Dun and Bradstreet was like all the AI providers, whether they're startups or big IT companies when data technology is their product, it's different. Second are there's certain industries and as a result, certain companies within those industries. Are digital natives. They were data savvy for a long time, long before AI came in the picture. They were always though, they were not selling data like Brad Street. It's not their product, but they were harnessing the power of data to get ahead. This is true for almost all financial services. Retail companies, telecom companies, they have, they had data readiness for a long time. The foundation was there and then they started building analytics on top of very common example could be if we compare Uber with taxi service, both take you from one place to another, but Uber found a way to harness data. Much more than the taxi companies either. So this is an exception in the transportation sector. But then there are the vast majority of companies who never thought of or never understood that data is such an important resource. But take all. Either got attracted by the hype of AI or just this idea, that AI could solve all problems and started there instead of thinking through whether it they have the groundwork or they had a fear of missing out. Everybody else is using ai, so I have to use AI too. These are the companies that the foundation is not there. And because you can't see the results of data investment right away, data has to go through. Analytics has to draw insight from that data. Then the company will have to have executives who are actually willing to implement the results, the things that are seeing in data science. And only after that the return on investment good fund. So those are all long-term long-term results and you don't see immediate gains. So they all got attracted by maybe AI could do something. And these are the companies where there is the issue of data readiness.

Ben parker:

Yeah, you see that?'Cause a lot of firms that invest more or it does feel like the market is just beginning to shift and there's a lot more focus on the data foundation. But like previously used to be like businesses, a lot more investment in hiring data scientists to build models.

Nipa:

Yeah, very often that's the first step in know hiring data scientists without creating the foundation. Then there's another thing. Often data scientists are hired, but they are not part of the problem solving teams in the folks who are used to dealing with data the traditional way. They give data scientists, data related tasks. if the data scientist really understood the business problem and was part of building the solutions, they will probably do different kinds of data science. They will ask for different kinds of data. They will process different kinds of data, and that would lead to different results.

Ben parker:

So is this. Where firms have not got this fully on board, is it more pressure from C-Suite or the board to show results from AI or is it hidden costs, et cetera? Skipping the foundation steps. I.

Nipa:

So in my view, you know what I have seen in most of the companies where the data readiness is not there. It's actually a misunderstanding of what data readiness means. There is pressure from C-Suite, but not always conviction from C-Suite. They're still in, show us what you can do mode and often the data issue is confused with technology issues. So most companies focus only on the technological component of the data problem, like you know how to store the data. Let's focus on cloud technology. So if you divide any problem in what and how the focus goes on the how, the technology aspects of it, but the other side of it, what will. How you get the data ready should be connected to the business problem you are trying to solve. What sort of analytics will you apply to the data? And as you get insight from the data, is there executive readiness to make real changes in the company? That's the only time you will get the results from data application of AI or data science. Many times these are still kind of experiments on the side and you don't get results till you actually apply.

Ben parker:

So why do you think there is a mis misunderstanding in leadership about what data readiness actually means? I.

Nipa:

I'm not sure I know the answer. Why? I think it'll take some time, even if you simply look at the org chart of companies, most companies have a CFO, A Chief Financial Officer, right? Because it's now established that you need a CFO to run a company. Similarly, you will see A CTO, a Chief Technology Officer who's again trying to formulate the how aspect of it. In very few companies, you see a CDO, chief Data Officer and even fewer, you see a Chief analytics officer very quickly. There are AI leaders coming up, and I honestly think all this hype about ai, especially the generative ai. Then ai, I think the. Most impact that will come from all of it is they will realize that to do those, really shiny things, you will need data readiness. And you will need an office that's responsible for it. I think it's the lack of that, too. To many executives, data is still Excel spreadsheet in it. They have bought into the hype, but not necessarily into the foundational aspects of creating this data environment.

Ben parker:

Have you ever had executives just come to you say, look, we've got lots of data so we are ready. Do you have face that situation?

Nipa:

All of the time. And that's the thing, ages ago problem was scarcity of data. How do we, that's the basis of statistics. You use small amount of data, you use a sample. And you draw conclusions from it. A conclusion that's expected to reflect the whole population, not just the sample. Then with advances in computer, it got turned on its head. Everybody has data. Way too much data. People don't know what we do with all that data. The definition of data change. There was a time when if somebody say data, you will think of rows and columns of numbers. Now data is in PDF files, in text form. Data is in voice recording data is in in videos. There is huge amount of data and that is often misunderstood as data readiness. So there's all that data. It's not in one place, it's in 10,000 disparate systems. There's duplication, there are inaccuracies, there is often no governance, no security. There is no record keeping which data was updated and which was not. As I was saying there aren't too many CDOs, but when they're there and when they exist in different, with different titles, their job is hard, really hard, and it's often more difficult to show results. So yes, that was the most common thing. Instead of explaining the business problem we had, we have all this data, what can we do?

Ben parker:

Okay. Interesting. Then, can you provide an example of why AI initiative would not take off then?

Nipa:

That's what happens. There is a desire to apply ai, but apply AI to what, if the foundation is not ready, then the conclusions reached by, let's say in machine learning algorithm something similar to that may not actually be reflecting the truth. And I think the problem at the other end is even more severe when a group of data scientists are coming up with a conclusion about something. And if implementation of that requires a major change in how the company operates, they need to often stays as interesting research but not implementable. So it's this whole journey. You can think of it as data collection, data storage, data management applying data science to it, and then implementation. If this whole chain is not established, then only the implementation will not work. You know there are AI vendors out there, right? They're trying very hard. They have their softwares and they have their platforms, and it's very tempting for non-data native companies to think we can license one of those softwares and that will solve all our problems. But that's not how it happens, to get your data ready for using those softwares is also a.

Ben parker:

Yeah, it's a lot, obviously. Yeah, a lot of work. This, it's always, it's not as easy as it sounds, as it I wish it would be, but that's obviously, it's part of business, isn't it? Is there. The lack of focus on data readiness as a resource issue, is it a skills gap or more a like cultural problem within an organization?

Nipa:

Perhaps a little bit of everything, most of the time what I have seen is it's a cultural problem to start with. That leads to the skills gap and the resource gap. If the data leader is not at the right table with other executives in a position to influence the CEO, who's not expected to be a data expert or a technology expert, then there is no one to attract the investment to hire the resources with the right skills. I think it has happened for technology. CTO is a much more established role in different organizations. Still that happens for data professionals. I think this is a hard problem to solve.

Ben parker:

Yeah, definitely. Again it's, because every business has its problems and it's, everyone's got their own objectives, isn't it? So there's no easy answer, unfortunately.

Nipa:

Yeah, and perhaps, personally for me, I have escaped or sidetracked this problem by working for companies that are consulting to others on data analytics. When that happens, it's your products. It's part of the revenue stream. But if you are internal support to the company and companies need a lot of data analytics support, that's when this lack of investment becomes a much bigger problem.

Ben parker:

So end then, what practical steps should companies take today if they want to become truly data ready in the next, say, the next 12 months?

Nipa:

Yeah. Perhaps this answered in many different ways. I'll follow something like this. I think the most fundamental thing that first has to be figured out is what will you do with the data when it is ready? That step has to be decided first. So as practical steps, perhaps the best approach is to form a COE a center of excellence, not really for data, but for data-driven decision. So this COE should include employees with data expertise in data science, expertise in technology, but also should include folks with. Domain knowledge and expertise in what the company does, and there needs to be a powerful C-suite sponsor for this COE. Then the COE should identify a handful of problems, some big, bold, audacious problems. Solving, which will, seal the deal. The company will be in, in a very different place if those problems are solved and some easier ones that can be solved within a, if the problems are all too complex, company will probably run out the peak. If they're too simple, then the benefit may not be obvious. Then from the very beginning, there needs to be some sort of agreement about what success looks like. As I was saying, it's an evolving area, so it could change over time, but there needs to be a tracking mechanism to figure out is it, is the program doing what it is supposed to. And once the business goals of the company and the data readiness project goals are aligned, they normally focus on data platforms, data collection management, data quality, security and as I was saying, data without data sciences is meaningless. So there will have to be a data science team. That's capable of building the solutions, participating in, in, in solution making business solutions, not just the data solutions. And they're also capable of making buy versus build decisions. There are so many data AI companies that are specializing in it there's no harm in using some of those solutions Lastly, I think it's the most important. There will have to be commitment to implement the things that comes outta this group or this project.

Ben parker:

Brilliant and really impressed with your insights today, nepa obviously amazing. You've managed large teams, hundred, 200 odd people. That's obviously impressive credentials, and it's been pleasure having you on the podcast.

Nipa:

Thank you, Ben. It was very nice talking to you.