00:00:00:00 - 00:00:37:05

Gowri

Have you ever looked at smart people and wondered, hey, what is their journey like? How did they get to this point? Where do they find inspiration? What influences them? Welcome to 1000 paths, a series of conversations with professionals from all walks of life. Here we explore how they made the decisions that shaped their life and career. The series is brought to you by data, by cloud and All-In-One Data Platform for fast growing businesses.

00:00:37:07 - 00:00:47:04

Gowri

Because in a time of AI and by what still fascinates us is the stories of humans.

00:00:47:06 - 00:01:13:12

Gowri

Today's guest is Abhinav Jahan, an analytics and strategy leader with over 15 years of experience across industries, banking, retail, travel and hospitality. He's also worked across geographies in India, in the US, the Middle East and Southeast Asia. He's eight years of interest include customer loyalty, customer experience, and marketing analytics. Let's hear from him today. Hey, Gowri. Hi, Abhinav.

00:01:13:14 - 00:01:35:17

Abhinav

As you said, my undergrad was in production and industrial engineering. Interestingly, it was a four year course. B.Tech. Like all of us do. in the fourth year, there was this B.Tech project that I had to do, and I got an opportunity to actually work on a live project that one of my professors was sort of consulting this, industry on.

00:01:35:19 - 00:02:01:23

Abhinav

So effectively what that was a back then and I, you know, use the word it was the decision support system for sort of procuring some, raw materials, etc., etc.. And that was the kind of B.Tech project that I ended up doing. so that was my sort of first brush with analytics effectively. I was using the most cutting edge stuff back then, which was sort of Excel and macros and linear solver.

00:02:01:23 - 00:02:27:19

Gowri

But, that actually gave me a little bit of an idea of, you know, what data is, how can we use it? So I think that was my sort of first brush with analytics. But right after engineering, I joined this technology consulting company where essentially we were developing tools or softwares effectively for hedge funds. Right. And a whole host of those softwares, for example.

00:02:27:19 - 00:02:54:21

Gowri

And I got, you know, exposure in technologies like SQL, like C, sharp coding technologies, etc., etc.. But what we interestingly, what we were doing was we were actually working for hedge funds who were effectively utilizing back then, we didn't call it analytics, but you would be utilizing this, you know, technology platforms, for example, for understanding the markets, understanding their own portfolios, etc., etc..

00:02:54:21 - 00:03:20:19

Gowri

So I think that effectively pushed me a little further, both on the technology side, making me comfortable around, some of these languages, but also, you know, just the use cases, etc., etc.. Moved on to Blackrock. again, very similar, but it was focused on pricing and, you know, valuations and things like that similar kind of thing. Again, we were not calling it analytics back then.

00:03:20:19 - 00:03:24:10

Abhinav

We were just valuation and pricing teams etc..

00:03:24:12 - 00:03:32:10

Gowri

Was that the vocabulary back then that it those valuation and pricing, you don’t bring in data or analytics at all into the.

00:03:32:12 - 00:03:35:12

Abhinav

Yeah. It was the valuation team effectively.

00:03:35:18 - 00:03:36:10

Gowri

Yeah.

00:03:36:12 - 00:03:46:15

Abhinav

But again look at I started looking at a lot of underlying data. Right. So this was complex. You know models that we had built looking at underlying data.

00:03:46:15 - 00:04:17:17

Abhinav

So and it was sort of evolving I think the industry was evolving as well. We had started using some basics as reporting tools, etc., etc. after that I started my own company in the same area. Right. So I created this company called analytics where we created this underlying platform for valuation, etc.. Well, so effective, like continuing what I was doing with Blackrock, but then started my own organization.

00:04:17:19 - 00:04:46:18

Abhinav

so I was there for about two years and the company was not where I effectively wanted it to. My, my co, partners continued with the company, but I decided to do my MBA. After that, I realized that we were all very similar in thinking, you know, quite from a technical perspective, etc., etc. but I think that diversity that, is often required to sort of bring, you know, company to life was lacking in the leadership as well.

00:04:46:20 - 00:05:09:18

Abhinav

so, yeah, did my MBA after my MBA, this was the time when I joined this, analytics consulting company and this was the proper customer, what we know as customer, analytics now. Right. So I work with a lot of banks for customer analytics, risk analytics, etc. got a really nice exposure in the kind of things that are sort of very prevalent right now.

00:05:09:22 - 00:05:33:00

Abhinav

I was primarily working in the Middle East and, Singapore, southeast, Southeast Asia region there. after that, I went back to India, and joined American Express again, very similar piece of work. So effectively, you would see that I got a an idea of working on analytics project as a consultant, but also as an internal team member, etc., etc..

00:05:33:00 - 00:05:39:16

Abhinav

So that's how sort of, the evolutions came through. And of course now I am with, with clients.

00:05:39:18 - 00:05:51:02

Gowri

One of the things that caught my eye about your profile was your work at the Amex Center of Excellence for analytics. What would be the purpose of such a center of excellence? Can you tell me a little bit about that?

00:05:51:04 - 00:06:19:14

Abhinav

Yeah. look, I think if we look back, this is not just, me and my experience, but if you look back, I think that there have been cycles of centralization and decentralization, particularly in the area. Right. So with, the analytics Center of Excellence, that I was working for American Express on effectively what that was was there was this central team doing similar kind of work across 27 international markets.

00:06:19:14 - 00:06:52:08

Abhinav

Right. So I think there was sort of some synergies around how we should look at the voice of customer across all the markets, etc., etc.. So I think there were definite sort of synergies that we can look at one versus the other. So I think that was the whole idea of bringing the center of excellence in the in the exciting bit, as I said, was looking at 27 international markets, there are things that we could learn, but, you know, see, a Sweden is very different from, you know, Japan, for example.

00:06:52:08 - 00:07:28:11

Abhinav

Right. So there were so those nuances that we'd still have to look at, etc., etc.. So I think that was the kind of motivating factor, I would say, for creating the center of excellence. There were still some certain decentralization. There might, might have been for for bigger set of markets like the US, etc. there were still analysts and analytics team that were present on shore, and we would work with them, closely understanding what we could bring to the table versus what they could do, etc.. So it's quite exciting in that regard.

00:07:28:13 - 00:07:45:03

Gowri

I think organizations have to reach a certain scale to set up a formal center of excellence. But is there anything that a founder or any leader trying to be more data driven within their organization could benefit from? What would what is your take on that?

00:07:45:03 - 00:07:51:17

Abhinav

yeah. Look, I think data culture generally, it just makes or breaks things, not just with American Express.

00:07:51:17 - 00:08:20:01

Abhinav

Right? I've talked to a lot of people, a lot of my sort of colleagues have looked at, projects, you hear about projects that are sort of going through. Some of them don't sort of succeed. Right. So, let's take an example. Right. So self-serve and being, you know, everything, everyone would be able to do everything, etc., etc. so much focus and you would think, that most of the projects undertaken would be success, but it's actually the contrary.

00:08:20:01 - 00:08:46:02

Abhinav

Right. So, about 60, 70 if not more percent of self-serve projects do not work. Right. But essentially what that is, is and this is my sort of firm belief, is it's not the tool that they used or it's not the, program that they use. Actually, it ends up being data culture. Right? So it's yeah, from my perspective, I'd say 2 or 3 things that are sort of, key.

00:08:46:02 - 00:09:08:08

Abhinav

Right. So I think data culture, while individual teams might want to sort of propagate it and it is necessary, I think one of the most important things I'd say it has to come from the top right. And this is absolutely true with the organizations that I have worked with as well as, you know, younger organizations, startups, etc., etc..

00:09:08:08 - 00:09:29:24

Abhinav

Right. So you have to understand at the top level that this is something that needs to have focus and sort of, agreement at the top level. Right? So I think that definitely is one. The second thing would be this is not something that you sort of do once and then forget about it. So you have to be persistent and consistent around this.

00:09:30:01 - 00:09:55:06

Abhinav

and then I think with that comes in a lot of sub things. Right. So we talk about data literacy training, etc., etc. within the different teams with our stakeholders and so on. So, yeah, I think it's a it's a program of sorts that has to be consistently sort of rolled through with the, you know, direct ownership and sponsorship from, from, the top.

00:09:55:08 - 00:10:31:11

Abhinav

The final thing that, I would say on this one is data culture is decision culture. Right? So as long as our analysis, that the teams are doing are being used for decisions, then it's only then that actually a culture that performs. Right. They could be individuals or teams that are doing fantastic work, in silos. But if they are not connected to, you know, decisions, for example, I think that bond, you know, is is the most important thing. So I think that is what I would say, you know, to just round it out.

00:10:31:16 - 00:10:45:02

Gowri

I think that's a great way to put it. Data culture is decision culture. What you do with the insights, the teams, everything sort of coming together. I think that makes all the difference. Since you mentioned that it has to be top down.

00:10:45:04 - 00:11:03:11

Gowri

I was wondering about startups and small organizations where the founders or one of the founders don't necessarily have a tech or data background. So what are some ways in which they can still empower themselves to be more data driven in the way they make decisions?

00:11:03:13 - 00:11:10:09

Abhinav

Look, increasingly, this is what is happening, right? That the technology should be at the back end, right?

00:11:10:13 - 00:11:40:08

Abhinav

It should never be that I'm using this fancy new product, etc., etc.. So I think it has to be business led. So just understanding what exactly are we trying to do? And then getting people who sort of understands and what exactly are we trying to do, just rowing in the same direction effectively? I think that is important. So sure, you might not individually understand what exactly the analytics teams might be doing, etc., etc. and it's it's both ways, right?

00:11:40:10 - 00:12:08:13

Abhinav

I think it is that I need to be actually there's this this debate that also comes in, which is you should you should we call it data driven, business. I think prefer data informed, as one of the things. So, what the difference is, you know. Yeah. Like, there, data informed means you would look at the insights that the team are sort of building.

00:12:08:13 - 00:12:30:11

Abhinav

You can still choose to use your sort of expedients, etc., etc., but I think you, the analytics team, actually have a seat at the table. Right. So, you know, hopefully that sort of answers your question in that you don't need to be a data scientist to sort of drive that data culture. You should know exactly what you're doing, which you will.

00:12:30:13 - 00:12:59:15

Abhinav

and, and I think the best is getting someone on the from the analytics sort of team on the table. You know, I think that is where everything sort of effectively starts. And you're seeing that across, enterprises as well. Right. So there was this, rise of this chief information officer, I don't know, ten, 15 years ago, but I think that is sort of, evolving, to a data, chief data officer or CDA or something of that nature as well.

00:12:59:15 - 00:13:08:10

Abhinav

Right. So I think getting a seat on the table and just, you know, accepting that, you know, this is sort of valuable to, to, to look at, I think

00:15:31:17 - 00:15:42:06

Gowri

Tell me about how you handled customer experience and built loyalty while at American Express, which is sort of a leader in the space.

00:15:42:08 - 00:16:03:03

Abhinav

Yeah. Look, I think one of the things within, within American Express that my team was responsible for, you would note that American Express has these sort of super premium cards, right, which are sort of your platinum flags and signature cards, etc., etc.. And what they get is, you know, those are the premium or the most premium of cards, right?

00:16:03:03 - 00:16:36:05

Gowri

And you will typically get relationship managers with that. So if I am, you know, signature card holder, for example, for Amex and I want to go to, I don't know, New York or something like that, I have the. So what I can do is I can actually give my relationship manager a call and I can tell them, I want to go to New York and I want to dine at this particular, five star Michelin restaurant, etc., etc. and then the relationship managers actually make that happen at the back end, right?

00:16:36:05 - 00:17:14:11

Abhinav

So as you said, American Express is absolutely known for NPS, right? So at all costs, the customer should be happy and satisfied. Yeah. So our team was responsible for understanding the performance for relationship managers, providing them with the right sort of incentives, measuring their performance and then actually also giving them prompts around what can be done, be done, etc., etc.. At the back end, there was a lot of data that was, that we had on our at our disposal and just utilizing that to sort of serve up as best as we can to the clients. That was what was it?

00:17:14:13 - 00:17:30:00

Gowri

Tell me about situations where you had very little data or no data at all, and you still had to go ahead and make big decisions. How have you tackled? How do you tackle such situations?

00:17:30:02 - 00:17:56:07

Abhinav

You would never find a person, an analytics sort of head, saying that I have enough data. Right. So I think, will always be data that you can actually use to infer some certain things. And I think that is where the kind of advances have taken place, right through machine learning, etc., etc. but also intuitively, you can make educated guesses looking at A, B and C, right?

00:17:56:07 - 00:18:16:11

Abhinav

And you know, in my, current, for example. Right. when we launch a new partner that in an industry that we've sort of not worked on, what do we do. Right. So one of the work that my team does is target the right set of people with the right sort of marketing offer, for example. Right. So that is sort of bread and butter.

00:18:16:17 - 00:18:35:23

Abhinav

What you typically would do is you would look at historical data. Okay. you know, I want to sell credit cards, for example. These are the type of people who've bought credit cards from us, you know, create a fantastic model, provide to an offer based on we feel might work, send out an email or an advertisement, etc., etc..

00:18:35:23 - 00:18:56:12

Abhinav

Right. So it works with the data that you might have. Now, if you are entering into a partnership with, with a company or in an industry which you've never worked on, so you'd effectively had, don't have any data, right? So, for example, if it is a fuel provider, right, a petrol station, etc., what would you do then?

00:18:56:12 - 00:19:23:11

Abhinav

Right? And we find ourselves in that situation, often always, etc.. So there's a whole host of things that you could do, right? The first one is just looking at doing market research and understanding what are the types of people who might be interested in this. If it's a high sort of value, you know, offering or it is a premium offering, better if it is exclusive, etc., etc., so you can make those sort of judgments and then sort of create personas, for example.

00:19:23:11 - 00:19:44:00

Abhinav

Right. The other thing could be looking at adjacent sort of areas, or maybe just do a survey, to 20,000 people, ask those kind of questions, etc., etc.. So effectively what I'm getting to is, what are what is the next best alternative tip that you can actually get to? It is something that we, that we generally go for.

00:19:44:02 - 00:20:08:03

Abhinav

And I think the more important thing in those situation is as soon as we do something, we have to, you know, do a fast follow and, you know, learn from what we are sort of seeing, etc., etc.. So I think that, loop back essentially becomes even more important. It has to be even faster, etc., etc.. So yeah, of course that absolutely is a problem.

00:20:08:03 - 00:20:30:22

Abhinav

But there are there are options that you could sort of do which are which we feel are adjacent and it would definitely depend on, you know, the business, that sort of understanding. So talking to the business side themselves and trying to understand what they think of it actually can give us clues around what we may be looking at in the data that we have.

00:20:30:24 - 00:20:43:21

Gowri

You spoke of the advantages of having a strong data informed culture. So what would you say are some of the factors, processes or behaviors that can help cultivate such a culture?

00:20:43:23 - 00:20:54:11

Abhinav

Look, I think the first thing by far has to be that and we've talked about this, right. has to be that acknowledgment from the top level that this is valuable, right?

00:20:54:11 - 00:20:54:18

Gowri

Yeah.

00:20:54:19 - 00:21:23:07

Abhinav

I think that is by far the most important thing. right. You would think that creating the best model is the most important thing, but I think it is just getting that acknowledgment that, yes, this particular team is valuable and we should listen to them. I think that is the most important. I think what goes hand-in-hand with that is, you know, once someone is showing confidence in this particular team, you know, the team has to sort of show value quickly.

00:21:23:11 - 00:21:45:06

Abhinav

I think that is important. And again, it will not it need not be hundreds of millions of dollars, you know, saved, for example, right away. But I think, to get that ball rolling. Right, I show confidence in you and I need to sort of get an immediate sort of a reward, effectively, for a lack of a better word, where we say, okay, yes, this is actually working and go and go on.

00:21:45:06 - 00:22:06:24

Abhinav

So I think what that sort of translates to is the use cases that we end up, you know, picking up. Right. So, do a quick review of what is the capability at the moment in terms of both the, the data, the technology, but also the people, and then come up with something that can be done relatively quickly.

00:22:06:24 - 00:22:24:15

Abhinav

And that is sort of valuable to the stakeholders of working with them. And then once we have gathered a couple of wins under our belt, you can get a bit more ambitious as we go along, I think. So I think it's also picking up. So the second point then is also picking up the right sort of use cases at the right time.

00:22:24:18 - 00:22:46:03

Gowri

Right? So if you're trying to sort, you know, creating a model to solve world hunger while you're just, you know, just started, I think that that might be a recipe for disaster, for example. And again, so the evolution, I would say would be would be one important thing. What else? I think the third thing is just keeping it simple.

00:22:46:05 - 00:23:07:17

Abhinav

Oftentimes I think, this is something that I've seen in, in, you know, in my industry, our industry effectively, of analytics, where, you know, the most complex thing is should be the best thing, but that, that really sort of happens. I was just reading somewhere. I was reading a book where it says, picture. I think Picasso, created this.

00:23:07:17 - 00:23:34:20

Abhinav

12 photos of a bull. Right. It's that Picasso. you know, what he did was the first picture was the most exquisite sort of, bull picture. And then he deconstructed the bull where the last picture was just, you know, four, five lines, etc.. So, yeah, effectively, stating that in order to get to simplicity, you'd have to go through that kind of process as well.

00:23:34:20 - 00:23:52:18

Abhinav

But I think, just ensuring that it is simple and it is explainable. the stakeholder sort of understand what exactly are we doing? I think helps in, in selling what we are trying to do. And then, yeah, I think those might be the three things that I'd say.

00:23:52:20 - 00:24:05:02

Gowri

Now that you've mentioned selling, I am wondering what role does storytelling play in the presentation of data to a larger group of stakeholders

00:24:05:04 - 00:24:28:16

Abhinav

oh everything? I think, it is so important, right? I've seen I've been with my teams, have done it, I have done it where, you know, I've just we've created this long list of charts, beautiful charts and very insightful if you look at it and it just falls off at times. Right. So you have to understand what the individual is trying to sort of solve for and then come up with that story for it to be sort of actionable.

00:24:28:16 - 00:24:58:21

Abhinav

Right? So, it is absolutely something that is important. Increasingly that is that has been the focus of the industry as well. Right? So maybe ten years ago, the focus was how do we create this beautiful chart the last four, five, six years? definitely. you know, focus where we're saying, how do we sort of, presented to, because, you know, an analysis like anything is, is only as important as how you've understood it.

00:24:58:23 - 00:25:34:06

Abhinav

Right? So, yeah, absolutely. important. it's a tricky one. It is it is difficult. Right. So it's not that you do this particular course and you're good. You become good at storytelling. I think the key there is to just constantly be in touch with the stakeholders, try to understand, really learn and understand from them what they are trying to solve for, what are the problems, etc., etc. and then sort of coming up with, with the solution over your career, the span of your career, how would you say the analytics industry has evolved?

00:25:34:08 - 00:25:38:08

Gowri

Do you want to talk about some of the inflection points that you have observed?

00:25:38:10 - 00:26:05:18

Abhinav

Look, I think the first one is I would say democratization or the wave towards it. Right. So, this is how I see the industry, right? The, the earliest analytics, you know, individuals, teams, etc. they were sort of very technical. They would go they, you know, lock themselves in a room and then they would code for like 32 hours to come up with one beautiful chart, etc., etc..

00:26:05:18 - 00:26:34:18

Abhinav

So these were individuals who were sort of very valuable, very specialized skills and then trying to come up with insights which are super helpful. at that particular point in time, the industry then sort of move towards your tableaus and your power BI is right. So effectively what that was one step towards the democratization of insights where it was still the analytics team was sort of creating your dashboards and things like that.

00:26:34:18 - 00:27:06:10

Abhinav

Right. But the business user now has the capability of selecting a drop down, slicing and dicing on their own, etc., etc.. So it moved the words. There was some, you know, movement towards the the stakeholders sort of understanding what it is, etc.. I think the third move, which has already sort of begun and it will sort of, you know, continue to go in that direction with some of these Genii kind of tools is, you know, I'll just ask what is happening, and then hopefully we get the answer.

00:27:06:12 - 00:27:29:19

Abhinav

Business, is sort of lagging behind, but I think that is that effectively things will be moving. Now, that is not to say that. And the analysts would not have to do anything. I think it's also the evolution there where things would be sort of changing. So previously we were sort of focusing on coding, you know, creating dashboards, etc., etc. I think I'd see more and more data quality, right?

00:27:29:19 - 00:27:50:23

Abhinav

If you're feeding to these giant models, it better sort of spit out the right answer, you know, and there has to be a right input, etc., etc.. So the kind of work that the analytics team would be doing might be sort of slightly different as we go along. So different, effectively evolution of different skills and things of this nature.

00:27:51:00 - 00:28:13:02

Abhinav

And then the third one, so I think over the last five, 6 or 7 years, there was this Cambrian explosion of tools. Right? So I don't know, every other day you would think at year you'd see that there was this startup that has gotten series B funding of to do A, B, and C, and I think that is what, they call the modern data stack.

00:28:13:02 - 00:28:41:01

Abhinav

Right. So you have maybe five different tools to do one specific thing. Right? And you know, that was due to the time, that we found ourselves in cheap sort of, you know, investments, etc., etc. but I think that has sort of that era has completed. So I think consolidation is what I will see will happen over the next five years in terms of tools. And that that has sort of implications around how companies, are using the tools

00:28:41:03 - 00:28:50:20

Gowri

For a young, fast growing, fast evolving sort of company. When would you say is the right time to start building a data team?

00:28:50:22 - 00:28:57:02

Abhinav

Oh, right from the start. I might be a little biased, right? But,

00:28:57:02 - 00:28:57:16

Gowri

yeah,

00:28:57:18 - 00:29:08:22

Abhinav

yeah. Look, I think, there are different, times where you can or the team could bring in different value, through the, through the life cycle.

00:29:08:22 - 00:29:39:12

Abhinav

Right. So, get a, you know, an enterprise like, the one that I'm working on or you'd see any, any organization, it has, you know, data engineers and, dashboard developers, machine learning, data engineers, data scientist, etc., etc. you don't need all of those kind of flavors of people, at the start, for example. Right. But, as soon as the company goes live, there is data that, has started.

00:29:39:12 - 00:30:05:07

Abhinav

So that's coming through. so I would say, the kind of work will be slightly different. And one individual, for example, or 1 or 2 people instead of, a jack of all trades can actually help you much further rather than a specialized sort of a person that. So and then there's of course, this, this life cycle of when do we need which individual, etc..

00:30:05:09 - 00:30:26:13

Abhinav

What do I look for? I think, the most important thing is actually the, the, the attitude. I would say, of solving things effectively. Right. So you could teach anyone a new kind of a tool in any way. There are so many tools. Chances are that not everyone would be well-versed with what you're sort of looking at, right?

00:30:26:13 - 00:30:51:12

Abhinav

I think just ensuring that who, whoever is sort of coming on board has a the knack of picking up things, but also is looking forward to sort of learning, etc.. I don't sort of look at individuals who are sort of super technical. I would rather have someone who's good with the tools, but also understands the business. So I think those are the two things I would say. I generally

00:30:51:14 - 00:30:58:24

Gowri

thank you for your time. Abhinav. are there any parting thoughts that you would like to leave us with?

00:30:59:01 - 00:31:27:14

Abhinav

Yeah. Look, I think what I have learned effectively is everything should start with the question. I think that is the most important. we find ourselves, you know, discussing whether we should use tool A and tool B, and of course there are nuances, etc., etc., but, I think 70% is, you know, if I can see that is understanding what are we trying to set aside for from a business perspective.

00:31:27:14 - 00:31:57:15

Abhinav

Right. Interpret it which tool to use, what models to use, what algorithm to use, etc., etc. are all great, but I think the the most incremental value is just understanding what, the what. And then also sort of to close the loop once I get the answer, this is how I would implement it. I think those are the two biggest things that I think we, you know, we should sort of effectively focus on.

00:31:57:17 - 00:32:22:24

Abhinav

And I don't think those get enough sort of focus. I think most of the discussions, most of the blogs would say, you know, Tableau versus Baba or, you know, Python versus R, etc., etc. we don't get a lot of discussions where we say, you know, this particular decision or that particular decision, etc.. So I think it's those that would give the biggest sort of bang for the buck.

00:32:23:01 - 00:32:31:19

Gowri

That was Well-Put. Thank you so much for sharing your insights. It was lovely chatting with you. Have a great day.

00:32:31:21 - 00:32:51:07

Gowri

Hello listeners, if you found this conversation interesting, do look us up on Datavicloud.com. Thats DATA VI cloud.com. You can also follow us on LinkedIn for more great perspectives about data analytics, business intelligence, and just making better decisions.

00:32:51:09 - 00:32:58:17

Gowri

See you next time.

‍