The halls of once great companies are lined with strategies that had no chance of materializing, simply because everyone was focused on the return, and nobody was focused on the investment. Is this the case with yours?
You may have Big Data ideas, but how will your strategy be Big Data capable? The difference between an idea and a strategy is capability. The halls of once great companies are lined with strategies that had no chance of materializing, simply because everyone was focused on the return, and nobody was focused on the investment. Once you've painted a picture of where your company will be in the next three to five years, and have some idea of how Big Data will fit into that picture, you must think about your investment in the key capabilities that will support your strategy; three general areas to consider are people, process, and everyone's favorite Big Data topic today—technology.The key thing to remember about Big Data, as it relates to your strategy, is this: only in rare cases will Big Data and analytics actually drive your strategy, so when considering the key capabilities that will support your strategy, Big Data will most likely play a supporting role—not a primary role. For instance, if you've decided to pursue a market-driven strategy, your primary focus should be building marketing capabilities. Of course, Big Data will play a significant role in the development of your marketing capability, but the objective is superior marketing capability with the aid of Big Data analytics: specifically not Big Data capability with a marketing orientation. Keeping this critical guideline in mind, let's explore what it means to be Big Data capable. A society of intellectuals
Finding good people to support your Big Data needs is by far your biggest concern. Of the three general categories of capability, people (not technology) are the most important. For the strategist interested in building Big Data competence, this should be concerning. As you may be aware, there is a dearth of talented mathematicians and data scientists in the workforce, and it doesn't look promising for the future—especially in the United States. You must aggressively recruit the best and brightest with attractive benefits and take equally aggressive measures to retain them. You should also consider embracing a more global and virtual organization, as top resources are likely to be located around the globe.
Although data scientists are obviously an important component to your capabilities, don't make the mistake of high-intellectual myopia. Culturally, data scientists fall into the engineering subclass, which means they create a very dangerous situation in your organization if there's no balance. Data scientists tend to be explorers with little concern for economics. And like their engineering superclass would suggest, they tend to ignore success for the sake of perfection. This can have real damaging effects on innovation, especially if you arbitrarily transition their analytic talent into product development talent. It may seem obvious, but you cannot rely on data scientists for advice on running your business unit—that's what managers are for. You must have leaders, managers, and/or coaches in place to make sure all this great analysis translates to economic results. The trick is to find good managers that can work effectively with analytic-minded talent. In truth, they're not very common, so this presents another significant challenge that you must take seriously.
Finally, you must of course have content experts. Again, this will completely depend on your strategic driving force. For a market-driven organization, you will need marketing experts. For a products-offered organization, you will need inventors, and product development specialists. Just like your data scientists, and your management team, you must strive to recruit and retain the very best content experts for your organization. This is not an area where you want to settle for what's available. If you can't find the right skills in the available workforce, recruit resources with a good base and a great attitude, and plan to cultivate these skills in-house.The return of SPC
The development and improvement of processes play a significant role in the capabilities the support a strategy that involves Big Data. The good news is that analytics and process control already have a synergistically effective history. You may have heard of Statistical Process Control (SPC), Total Quality Management (TQM), and/or Six Sigma which all have their foundation in statistics applied to the development and improvement of processes. Fortunately, the same resources that build your core Big Data competence can also be instrumental in constructing and optimizing the very processes that support the competence.
Process control is important in Big Data discovery, because it enforces consistency, reliability, and structure, in an otherwise amorphous environment. Developing a good discovery process requires a good balance between unbounded exploration and perfunctory bureaucracy. A good process succeeds at keeping the data scientists energized with challenging puzzles to solve, while concomitantly keeping the leaders energized with competitive products to sell and marketing insights to exploit. This is easier said than done; however, within the constructs of a good process control paradigm like Six Sigma, your chances increase immensely. When I studied to become a Six Sigma Black Belt at Motorola, I learned very advanced statistical techniques for developing and improving processes. This is level of rigor I suggest you employ in the development of your Big Data processes. When you have great people operating within a tightly controlled process that serves the explicit purposes of Big Data exploration, you've won the competitive battle. Technology just becomes an academic exercise at that point.
Being Big Data capable means having not only the vision, but also the substance to bring your strategy into reality. That means having the people, processes, and technologies that are beyond exceptional—you must commit to world-class Big Data capabilities to outpace the competition. For instance, Yahoo! has over 100,000 CPUs across 40,000 computers running Big Data technology—now that's commitment! However, much more important than technology is people, and then processes. Invest aggressively in the right people, and development of the right processes, and you'll be capable of almost anything.