Building An Analytics-Centric Organization
When I left the MIT Sloan School of Management in 1993, little did I know it at the time, but I was armed with the knowledge and mindset of a paradigm that would become known as Big Data. For the record, I owe tremendous credit to professors at MIT such as Steve Graves and Erik Brynjolfsson whose insights and perspectives left an indelible mark on my thinking. Although I could not fully articulate all of the details of what Big Data meant at the time, I did have an unwavering conviction that the lifeblood of a high-velocity organization would eventually be based on real-time analytics of and execution on mass quantities of data. And, to take it one step further, Big Data would revolutionize how companies operated and would drive the creation of whole new businesses (and perhaps even industries).
When I left MIT in 1993 for Intel, anyone could clearly see that computational power was advancing at a staggering rate and that the World Wide Web along with the broader infrastructure of the Internet were going to play a central role in this revolution. I knew that whatever “it” was – was going to be very big. But in all honesty, in my wildest dreams, I could never have predicted how big this digital transformation would become and how it would continue to evolve.
One afternoon in the mid-90’s, I wrote down a vision of an operating model for a corporation that would employ vast amounts of data and automated algorithms to enable the company to autonomously optimize and run its operations in real-time: integrating across operational domains that encompassed day-to-day tactical decision-making, operational planning and management and the long-term strategic decision-making. At the essence of this vision was an organization whose very existence and operational ethos was built on data and analytics. This is about as close as I ever got to a crystal ball. Imagine my delight and amazement when a few short years later, I had the opportunity to join a small, but growing company (Amazon.com), led by a visionary leader (Jeff Bezos) that shared this vision almost to a tee! I had found myself in the business and career analog of the “kid in the candy store”! Candidly, I thought that life could never be better – until I found myself a decade ladder in an even bigger role at Google.
During my entire professional career, I have built analytics functions and groups that operated at the core of any operation that I’ve managed. Fortunately, along the way, I have had a disproportionate share of success, but have made some mistakes as well. I’d like to share some of those key learnings along the way. This article will be mostly retrospective in nature – I’ll save the predictions for a future article.
If there’s one thing that I want you to take away from this article, it’s the following: start the transformation now! Dive into the deep end and get moving; you have no time to waste! Your competitors are not sitting still.
Make no mistake about it, at its essence, this is just another transformation for your company. That said, it is a transformation that is all-encompassing. Too many business executives view the analytics transformation too narrowly: They tend to view it as centering on tools or being driven by the hiring of “big data” analysts or machine learning expertise. Almost without exception, every successful analytics transformation that I’ve seen or experienced has started at the top – this transformation is as much, perhaps, even more, driven by a cultural change as it is by your hiring of new resources and technical expertise. Even today, however, too many business leaders and executives tend to be intimidated by analytics and math. My guidance to you (and you know who you are) is to get educated – fast. That does not mean that you have to get the equivalent of a Masters Degree in operations research or machine learning. It does mean seeking out an expert who can frame these capabilities in the language and the vernacular of a senior executive. This is the most perilous part of the journey because too many senior executives delegate this transformation to lower levels of the organization where it gets lost in a sea of other priorities.
The transformation cannot be haphazard. Like all good transformations, it’s built around a plan with measurable goals, timelines, and dedicated resources. This transformation is too important to be relegated to a side project. One word about goals – they have to be aspirational but realistic. There are too many wildly optimistic (and over-hyped) stories in the press today that are simply unrealistic and unbelievable when it comes to the promise of analytics and particularly machine learning and artificial intelligence. Don’t get sucked into the hype cycle. Truth be told, most companies are still early in their evolution around data analytics: be aspirational but realistic. Remember, companies like Google and Amazon have more than a decade headstart on this journey, and your company is not going to evolve to this capability overnight.
Okay, let’s dig deeper into the intricacies and nuances of data analytics. Consider the following as some of the key lessons learned from my 25+ year career in data and analytics:
- Always have an objective function in mind (e.g. know what problem you’re solving from the start): An objective function is the top level problem that you’re trying to solve. As an engineer, I would refer to this is using a first-principles approach to problem-solving: have a clearly articulated problem and goal stated before embarking on a path to solve it. This may seem self-evident or trivial, but you would be shocked at the number of times that I’ve walked into meetings where a solution was being developed before anybody could answer the following basic question: “What problem are we trying to solve?” In fact, I use this simple question: “What problem are we trying to solve?” almost daily. Start with the most important objective function and go from there – don’t try to solve every problem at once.
- Data is the foundation of a world-class analytics organization: Just as your house requires a solid foundation, so does your data-driven organization. And, unfortunately, this is part of the organization that is usually the messiest. Most of our organization’s information is siloed; resides in hard to query legacy systems or is poor in its quality (or all the above). Your organization is likely going to require a methodical process and effort to clean up and organize its data – and don’t underestimate the resources and tenacity that are required to do this well. One word of caution here – don’t get enamored with correcting every data sin of the past. Use your objective function as your guide – start small and only clean-up the data that is necessary to answer the fundamental questions framed by the objective function. Don’t let your organization get bogged down in a quagmire of data-clean-up where this effort consumes all of the energy of the organization.
- You can have too much data: All of our organizations (yes, even a company like Google) are victims of the double-edged sword of the data-rich; insight poor syndrome. Given our vast (and cheap) computational capabilities, we opt to collect all kinds of information – no matter how mundane it seems. Better to err on side of collecting everything rather than run the risk of losing out on some key insight! Ugh – what this means is that most of us are searching for the proverbial needle in the haystack when it comes to insight. It’s ironic that the very capability that makes all of this insight possible, complicates this effort to the nth degree. Again, be mindful of your objective function and stay focused on the data that provides the most insight into solving that particular problem. I know, abundant data can be addicting, but remember that data is the means and NOT the ends.
- Hire a world-class analytics leader: Every single analytics leader that I have hired had the following attributes: (1) They had a PhD from a top engineering or business school; (2) Their humility was inversely proportional to their intellect – they did not intimidate the organization with their brilliance, but brought it together; (3) They were bi-lingual. By that I mean that they were equally versed in the language of operations research and statistics (or other analytics approaches) as they were in an income statement, balance sheet or accounting principles; (4) They were a teacher – they mentored the senior leaders, mentored their peer functions and mentored their teams; (5) They were pragmatic and were not wonks who wanted to stay in the back-office or the ivory tower and write academic papers. They got their hands dirty with real-world problems and they loved this part of the job; (6) They possessed a knack for being about to look around the corners and were true change-agents for the organization – they were an instrumental part of the transformation process and took this part of the role seriously. In other words, they were strategic and innovative. My guidance: Invest a lot of time finding the right leader – this can be the difference between success and failure of your transformation effort.
- Build the right analytics organization: Likely there is not a one-size-fits-all model for this one. That said, I’ve had the most success in building a centralized analytics organization that acted as a central resource for the overall business functions. In this model, you can hire and retain the best talent and cycle them through the most interesting and challenging problems in the organization. The centralized resources also can act as stewards and teachers for the best practices of data analytics across the organization. Of course, this requires you to be mindful of the skills and personalities that you bring onto the central team. In other words, there is no room for prima donnas and people but don’t want to get their hands dirty. Oh, and it’s a war for talent out there right now – expect to pay top dollar for your analytics team. Don’t be stingy here!
- Know what tool (or analytical approach) to match with the appropriate problem: Data analytics and data science have existed formally since before World War II – it was just called statistics and operations research. With the advent of vast amounts of computational power, data science got rebranded as Big Data and with advances in neural net chips and algorithms, machine learning, deep learning entered into our vernacular. Be wary of the hype cycle – there is a myriad of data analytics tools and approaches that are well suited for one problem and not another. Machine Learning, Deep Learning, Operations Research, and Statistics (just to name a few) are not interchangeable. Everyone in the organization should have some knowledge as to how (and when) to apply a particular tool or technique to a class of problems.
- Get started – now! Did I mention that you and your organizations have no time to lose!