The Power of Systemized Decision-Making

Megan - 00:00:18: Today, my guest is Dhiraj Rajaram. Dhiraj is the founder and chairman of Mu Sigma, the world's largest pure play analytics company. Currently, Mu Sigma works with over 140 Fortune 500 clients. Dhiraj is responsible for the company's vision and strategic direction. As an entrepreneur at heart, Dhiraj is passionate about systemizing the art and science of problem solving to be done at scale. Before Mu Sigma, Dhiraj advised senior executives across a variety of verticals as a strategy consultant at Booz Allen Hamilton and PricewaterhouseCoopers. Dhiraj holds an MBA from the University of Chicago. Dhiraj thank you very much for being my guest on today's episode of CFO Weekly.

Dhiraj - 00:01:36: Hi, Megan. How are you doing today? Thanks for having me.

Megan - 00:01:39: Yeah, I'm great. And I'm excited about this topic. Today, we're going to be discussing the financial benefits of systemizing decision making, as well as the art of problem solving framework. We'll also be talking about the internal and external data to use within this framework and why it is that finance organizations don't always prioritize decision making. And finally, we'll take a look at some actual case studies. Dhiraj I'm looking forward to learning about you and from you. So let's get started.

Dhiraj - 00:02:08: Great.

Megan - 00:02:09: First, let's start with just kind of a brief overview of your career and your professional journey to date so that the audience has some idea of your background.

Dhiraj - 00:02:18: So I started off my life as a programmer working for PricewaterhouseCoopers, which became IBM. Went on to get my MBA in finance and strategy at University of Chicago, the Booth School of Business, and then worked at Booz Allen Hamilton as the strategy consultant. Post that, and actually while at Booz Allen, came up with the idea of helping organizations change the way they think about problem solving and built on the use of data to a large extent. And this was back in 2004 when there was no word called big data. There was a word called analytics, but not very popular. So we kind of went on to become one of the initial companies in this space of advanced analytics which later became big data and machine learning, and now it's called AI and all of those things. But the broad theme around all of this is that we are in the business of helping people make better decisions. And a good decision is one which is more true and is and happens. And the time to truth is lesser and lesser and lesser. So it's an optimization between the big T, truth, and the small t, time. So that's where we come in. We work with more than one third of the Fortune 500 companies. We went on to be India's first unicorn, one of the India's first unicorns and the first unicorn in this space. In 2022, we were the third most profitable unicorn in India. We took about $14 million of primary capital and recently returned about $900 million to our investors. So it's been a good ride.

Megan - 00:03:56: Yeah, an incredible journey for sure.

Dhiraj - 00:03:59: Yeah, and I would never expect it to be on such a journey. But more and more I am on such journeys, I feel like the fact that things happen to you and you make things happen are both concurrently true. So that's my background. Looking forward to engaging with you on questions that could be relevant for CFOs.

Megan - 00:04:17: Yeah, absolutely. And you've been the recipient of many accolades over the course of your career. But what recognition have you received that's meant the most to you, perhaps from a perspective that advanced your reputation or was vindication for some work that you were very proud of?

Dhiraj - 00:04:36: There's nothing that can beat recognition from my alma mater at University of Chicago. So I won a Distinguished Alumni Award, and that's something I treasure pretty much more than anything else.

Megan - 00:04:48: Congratulations on that. That's amazing. And you've explained before that the name Mu Sigma is related to your pursuit of meaning in a world that's drowning in data. And also that wanting to understand the unknown is another belief of yours. So in the world of data analytics, where are those great areas of unknowns that still fascinate you today?

Dhiraj - 00:05:14: The name comes from this perspective that Mu stands for expectations and Sigma stands for variance from those expectations. It could be surprise or disappointment. So if you think of all of the insights that come to us, all of the decisions that we take, it's an interplay between Mu and Sigma. So that was the thinking behind the name. And what I find today in large organizations is that these organizations have become very, very, very complex. And when I say complex, what I mean is there is a spaghetification of the organization. The purpose of any company is to take any idea and make it real. So operationalize these ideas. And as these ideas become real, they go from content in somebody's head to conversations to the compute environment, because everything is software right now. And eventually it is consumed by business and commercialized. But the uninterested complexity in our lives come in the way of this journey from content to conversation to compute. There's a lack of transparency of all the content. There's a lack of persistence of the conversations. And eventually, there is a lack of cumulativeness of the work that we do in the compute environment. We land up doing anecdotal work. The role of a finance department in many of these organizations is to be a traffic police of ideas. You want the good ideas to move faster and slow down the bad ideas. That's what you want to do. And if the finance department is not closely integrated in the decision sciences world, in the data and analytics world, it's going to be more and more difficult for the finance organization to make that happen. So our perspective was that you've got to be more data-driven. You've got to help create a transparency of all the content. You've got to create persistence in conversations in the organization. You've got to create a cumulativeness of the work, move from projects to programs. And for all of that, you need a new way of thinking about problem solving. You need a new kind of kitchen. And that's where we came up with changing the way one thinks about problem solving, moving from problem solving as entities to problem solving as interactions. And we built an entire suite of things, which includes artifacts and process platforms and people. All coming together. And that's what we call our art of problem solving ecosystem, which we use in many large companies. We have large software companies, retailers, airlines, pharma companies, across a variety of functions. We are helping them change the way they think about problem solving.

Megan - 00:07:59: And in the work you've done, just curious, do you find that organizations are still mostly siloed or that those siloed functions are starting to break down?

Dhiraj - 00:08:11: See, most of these organizations were built at a time which was far more deterministic than today. And in such a world, they pretty much were built to be effective by being very, very efficient in their execution. So what they did was they specialized into functions. Marketing did its thing. Finance did its thing. Supply chain did its thing. And they got really, really good at that. And that's how they competed to become good and operating on the efficient frontier of those functions. Now, as the world is moving from a very, very deterministic world to a very probabilistic world, which has a lot more uncertainty today, it's not enough if you're good at execution. You've got to be good at exploration too. You've got to be a producer of optionality. A world of uncertainty requires an organization to be producers of optionality. Now, the things that make you good at exploration are very different from the things that make you good at execution. Execution requires you to be very efficient. Exploration, you have to be cautiously inefficient and approach things from multiple dimensions. Execution requires you to have a very goal-oriented perspective to problem solving. Exploration requires you to have a journey-oriented perspective towards problem solving, where you are seeing multiple things, and then deciding which is the best place to be. So this means that while execution world was built on being vertical as silos, the exploration world requires you to be cross silos and therefore necessarily horizontal. So, horizontalization of the world is actually happening for this purpose. I just want to give the background of why horizontalization is important. But having said that, we are deeply entrenched into our hierarchies, into our organizational structures. So, it's very difficult for these organizations to devolve those hierarchies and organizational structures and become horizontal suddenly. I see a world, Megan, which is growing from organized hierarchies to networks. All work is going to become network. Your hotel industry is now a network called Airbnb. Your taxi company is a network called Uber. Your retailer operates as networks in marketplaces. So, everything is becoming networks. So, should an organization become more and more networks where problem solving is an interplay between product and marketing and sales and maybe finance and all of these people cross-functional coming together. But for that to happen, moving from a silo-based vertical structure, organized hierarchies to networks will require a whole bunch of changes in tool set and skill set, but most importantly, mindset. So that's where we come in and we are helping them build those frameworks and tools that work in those hierarchies, but still give them the effect of networks. And that's what we are trying to do. And be kind of a transformation Sherpa in that process. The transformation is obviously our customers' transformation. Our clients are the ones who are transforming, but we are kind of helping them with the client as a Sherpa. So that's what we do.

Megan - 00:11:41: Yeah, that's fascinating. So when you started Mu Sigma, and maybe you've answered this in part, but what was it about the process of decision-making that seemed so obviously flawed to you? And can you think of any real-life examples that highlighted those flaws?

Dhiraj - 00:12:00: So the dots always connect when you look backwards. But at that point of time, I didn't think that that's where my world was going to be. So now with the benefit of hindsight, I would tell you that University of Chicago as a school is built on efficient markets, rational thinking on one hand. But at the same time, it also espouses the opposite. Richard Taylor talking about irrationality, Kahneman and Tversky's thinking fast and thinking slow being introduced to us. So you had the finance guys, the corporate finance guys talking about rational markets to us. But at the same time, the managerial decision-making guys talking about irrational ways in which people are operating. So this seed of the duality of opposites was seeded to me very early on at University of Chicago. But if I were to look at where duality originated for me, it's obviously in India, which practices duality and thinking in opposites forever, all the way from its religion to its paradoxes in society, its diversity in the number of languages that get spoken here. So all of those kind of things are things that kind of seeded my mind in a world where I felt that, look, if organizations have to evolve, they have to find the organism in them. They have to find the living soul in them. And for that, you need the organizations to have very good feedback loops and very good learning cultures and evolving very rapidly. I felt the future of an organization is one wherein learning was going to become more important than knowing. Experimentation was going to become more important than experts. And in such a world, if an organization was built on keeping secrets outward, it will land up practicing it so well that it will even keep secrets within itself. And one group will not talk to another group and so on and so forth. So this was very evident to me when I was working as a management consultant at Booz Allen and I could see these inefficiencies in large organizations. And felt that the Mu IP to be effective is not intellectual property, but rather interaction property. So making these things happen meant that you had to create a science around this. Move the art of decision making to a science of decision making and move the science of decision making to engineering of decision making. And that's what we've been doing now for more than two decades, making this journey happen for large Fortune 500 companies.

Megan - 00:14:34: And when you think about hiring a decision scientist to work for Mu Sigma, what type of person and what experience do these scientists tend to have that make them suited to being decision experts?

Dhiraj - 00:14:50: The first and foremost thing that we look for is curiosity. Once we find that the person has a fantastic curiosity quotient, then we look at the person's ability to connect the dots and come up with insights. An analytical mind that can connect the dots, come up with insights. But that's not enough. You got to also have a mind which is oriented to the opposite of analytics, which is synthesis, where the person can articulate a whole bunch of disparate information into a concise story very quickly. Because people make decisions based on the stories that they hear. So the ability to communicate in a pithy format is going to become very, very important. And last but not the least, our business is a bunch of trial and error. And you've got to have a lot of grit to go through triangulation, multiple approaches before you get to a point where you feel you're confident enough to make a decision. So I think the aspect of having a growth mindset over talent mindset is going to become very, very important. So that's what we look for. And very early on, we can realize that these kind of people are not available out in the industry. So we got to not just focus on hiring some of these guys, but also making them. So we landed up building kind of a university-like ecosystem called Mu Sigma University. Inside our organization and hiring very, very young people for intellectual horsepower and then putting them through a rigorous curriculum, starting from design thinking and structured thinking to the basics of math, statistics, econometrics, operations research, to aspects of data engineering, data science, and then eventually operationalizing all of this with programming and engineering. And last but not the least, we also had to introduce them to a mini MBA so that they understand the basics of every industry because anything and everything that we do cannot be purely technical because it is very applied to either the marketing department of a retailer or the sales department of a pharma company, so on and so forth. So they have to be able to talk that language, if you know what I mean. So once people go through this kind of training ecosystem, then we practice learning by doing quite a bit. And we have an apparent three levels in the organizations. The first level is the trainee decision scientists and decision scientists, which operate as a network. The second level is apprentice leaders. And the third level is the leadership team. So the company's 4,000 people are just in these three networks. It's a very flat ecosystem. And we kind of are in the business of learning really, really quickly and making things happen. The one thing where we differentiate ourselves more than anybody else is how quickly we learn about something new in an organization. So we have systems and processes put in place to enable learning to happen very fast.

Megan - 00:17:48: And let's talk about why it's financially beneficial to systemize decision making. And does the process of systemizing actually take away from the human skill or the art of decision making?

Dhiraj - 00:18:01: Actually, no. I see if you look at anything and everything in the world, it always starts as an art. And that's when it's extremely beautiful and it becomes known for its expression. To it, becoming science becomes very explainable. Once things become explainable, it's science. And many people can do this just because it's explained to many people now. And then you want to scale it even more. It has to become engineering. And so the movement from art to science to engineering is a very natural movement in everything. Now, when you make things go from art to science and science to engineering, the advantage that you have is that the things that need to be art can continue to be art because you have given it a lot more time. And the things which can be automated is automated as much as possible. So basically, you get more time, relatively more time for art. And that's relatively more time for creativity. So if you look at a Michelin three-star restaurant, you will see that there's various levels at which the chefs operate. Maybe the big chef who is kind of dreaming up the next generation of food. But at the same time, you have a whole bunch of people who specialize in cutting the vegetables, curing the meat and this and that and everything. So it's kind of he's got his ability to automate things at various levels, using people and processes and maybe even devices to build an industrial kitchen so that he can focus on his creativity. Similarly, in large organizations, which are preparing themselves for a world of algorithms, they need to ask themselves, what does it mean to be prepared for a world of algorithms? An algorithm, when you're in the world of algorithms, all the insights that are happening are actually algorithms. And that algorithm is a network. Now, you've got to ask yourself if an insight is a network, then how should learning happen and how should knowledge exist? Is learning going to happen using individual questions or question networks? And is knowledge going to manifest as databases or knowledge graphs? So when knowledge graphs, so for a world of algorithms, you need knowledge to be networks as knowledge graphs and learning to be question networks. So those are all things that are going to be required for a world of algorithms, which means that a world of algorithms is going to require you to change your kitchen because you're going to make new kinds of food. So that's why systematizing is going to become very, very important. And new systems have to be created in line with everything else that's happening. So that's where we come and help people in evolving into these new systems, because we are seeing patterns across multiple industries going through the same journey.

Megan - 00:20:50: Yeah. And I know that that journey is not an easy one and that a lot of people don't even know where to start.

Dhiraj - 00:20:59: Yeah, but also I must tell you, many organizations have made a lot of progress. It's not going to be a linear, straightforward journey. It's going to be something where you'll have to get a few things right, make some mistakes, but keep moving forward.

Megan - 00:21:15: And you've mentioned learning over knowing and extreme experimentation over experts. So I'm just curious how CFOs can embrace these principles to foster a culture of innovation and continuous improvement in their own organizations.

Dhiraj - 00:21:32: Well, I personally think that a CFO needs to have a decision science group within his organization. And also, I would say that the chief data officer or the chief analytics officer should either report to the COO or the CFO. It depends on how activist the finance department in the organization is. If you're a company like GE, which used to, at least a long time back, used to have a very, very activist finance group, then it's not a bad idea to have the analytics department in the finance. But I think the chief data and analytics officer has to work very, very closely with this vehicle. So that's how I see it.

Megan - 00:22:14: And you've also mentioned that Mu Sigma supports one third of the Fortune 500 companies. Can you share some success stories from the work that you've done with these companies or some case studies that demonstrate where Mu Sigma has helped CFOs and financial decision makers achieve tangible results through data-driven strategies?

Dhiraj - 00:22:37: One example, I mean, look, we work across many functions in large Fortune 500 companies. But from a CFO perspective, one of the things that we kind of worked quite a bit on is financial planning and all of the analysis and analytics and mechanics that we have to create to shorten the financial planning cycle. And also creating a lot more optionality with real options in financial planning so that when something happens, which is not the norm. Like, for example, we just had a supply chain disruption because a bridge collapsed in Baltimore or the ship industry having a sudden shortage from an auto industry perspective. All of these are like war that happens in Ukraine. So any of these things could result in global supply chains being affected and financial planning will have to quickly respond to this. And all of this means that your financial planning cycle better be shorter and shorter and shorter. So these are kind of things that we help organizations with lots of analytics and simulations are a very key important tool in making all of this happen.

Megan - 00:23:45: And organizations today are drowning in data. So which sources of internal and external data do you and your decision-making framework prioritize?

Dhiraj - 00:23:56: That's a very difficult question to answer. There's no right answer to that. I can only give you a broad perspective that I have, which is that the three kinds of data, one is data that you have easy access that you want to be able to use, data that is not inside your organization, but you can scrape and get it from public ecosystems. But the most interesting data, which creates a competitive advantage for you, is data that you create. And you create for the purpose of a problem space or a hypothesis that you are looking to validate. Because those are the kinds of insights which others will not be thinking of, because that data is not easily available. So you know the joke about looking for the keys right under the lamp because there's light there. You don't want to be in that trap and just look at data that's easily available. So that's how I would look at patterns. But I think you're really right about the fact that the noise to signal ratio is something that we all need to be thinking about quite a bit, because there's so much data that sometimes you could be literally drowning in it. You know, we had a large bank who was going through serious challenges with their regulatory reporting and working with regulators. And literally every day, a report that goes, which has literally a million fields in it, and I'm not even exaggerating, it could be like a million things that could go up and down. And even if one of those numbers are incorrect, the regulator is going to categorize that whole report as something which is not meeting his requirements. This is an example of drowning in data. Because the regulator is going by a rule. The bank is stuck with the regulator going by that rule. And so much is lost in this process of following a rule which is not necessarily the smartest rule in the world.

Megan - 00:25:46: And when you think about the role of CFO in today's organizations, how important is it for them to be embracing and accelerating change? I know in years past, that role might have been more about stability and steady growth. But nowadays, it seems that if you're not harnessing new data and new technologies, that you're falling into irrelevance pretty quickly.

Dhiraj - 00:26:09: I want to bring this concept of dynamic stable structures to you. So if you're cycling, your chances that you are going to be in balance is far higher when you're moving than when you're still. And that's quite intuitive to you because when you're moving, the balancing becomes easier because your wheels act as gyroscopes and so on and so forth. Similar to that is in a world of constant change, a CFO who is focused on stability is actually running the risk of going into imbalance a lot more. He better be dynamic because you cannot just operate for resilience. You have to operate for anti-fragility, which means that it's not just about not breaking because of change. It's actually about getting better because of change. And for that to happen, the CFO has to be a strategic partner to the CEO and be responsible for making sure that the CEO has many more options in his quiver to use and make things happened.

Megan - 00:27:17: And with the rapid state of technological changes, how has your decision-making framework changed in the very recent past to account for things like AI and machine learning? Are you always trying to adapt your framework or do the principles remain the same?

Dhiraj - 00:27:35: There are some parts, our values remain the same, whereas our strategy is constantly changing and our tactics are changing faster and faster and faster. So if you were to think of values, strategy, and tactics, values remain the same on one hand, but tactics are constantly changing. So I think the perspective I would say is just to be a little bit more precise in that answer is I would say in today's world of algorithms, we have to orient ourselves towards knowledge graphs and creating ontologies of how data is stored. And the second thing I would say is that you've got to work on solving specific problems and catering to the urgent needs, but you also have to work on creating a better art of problem solving and building that kitchen. So while you're making new kinds of food, you have to constantly work on building a new kind of kitchen too. So both of those things have to work in parallel with each other. So you're literally kind of changing the, improving the engine as you're driving the car. So that's something which is hard, but has to be done.

Megan - 00:28:41: And last question, in today's turbulent world, what is it that keeps you up at night as you think about the future?

Dhiraj - 00:28:50: The thing that keeps me up is the breakdown of social structures in a world of AI. You're going to see more and more people not belonging to what I would call the useless class. And many of these jobs will need lesser and lesser and lesser people. As that happens, our social structures have to evolve very, very quickly to support such a world where the Pareto is going from 80-20 to 98-2. More and more wealth is going to accumulate with less and less people. A world of information is going to mean that you're going to have more and more polarization. The left is going to become more left. The right is going to become more right. And this is a common pattern that's going to happen across every country. And the US is going to be the first complete information economy. So it's going to face before everybody else. But after it faces it, others are going to face it too. So you're going to see things like polarization happen everywhere. But I don't think we are ready for those social structures to face all of that. So that's something that keeps me awake. And the second thing from a micro perspective, what keeps me awake is that organizations are still stuck in world of hierarchies instead of moving very, very quickly to a world of networks. So I see that many of them are going to be disrupted just because of the fact that they have not moved fast enough to a world of networks.

Megan - 00:30:14: Dhiraj, thank you very much for being my guest today. This has been a fascinating conversation.

Dhiraj - 00:30:19: Thanks, Megan. I appreciate you having me with you.

Megan - 00:30:23: Yeah, I really enjoyed speaking with you. And thank you for finding the time to be here with us today. I wish you and Mu Sigma, all the best as you tackle all of the problems that you're tackling.

Dhiraj - 00:30:35: Thanks again.

Megan - 00:30:36: To all of our listeners, please tune in next week. And until then, take care.