The organization that wants to achieve causal ambiguity must focus on big data innovation, retaining key analytic talent, and protecting the organization's information prowess.
Good magicians learn amazing illusions to entertain their audiences—great magicians invent them. Houdini, one of Americas most celebrated escape artists, started his career by freeing himself from handcuffs, ropes, chains and straight-jackets. As others witnessed his rise to fame, they began to imitate his feats.
In response, Houdini invented the more elaborate escapes which form his legacy, such as: The Milk Can Escape, The Overboard Box Escape, and The Chinese Water Torture Cell. In many respects, building a Big Data strategy is similar to becoming a master magician. To compete with big data, you must ensure nobody can duplicate your tricks.
There's an important concept in strategic management theory called causal ambiguity. As its name implies, causal ambiguity happens when competitors, new entrants, and possible substitutes cannot pinpoint the cause of your market performance—they cannot reverse engineer your success because they cannot figure out what's causing it.
Developing your organization's big data capabilities is a very progressive way to accomplish this; however, it takes a long-term perspective that transcends any shorter-term strategic goals. The organization that wishes to achieve this distinction must focus on big data innovation, retaining key analytic talent, and protecting the organization's information prowess.
Big Data R&D
Superior innovation with big data involves inventing new ways to analyze data. There are plenty of mathematical models that are freely available in the public domain. These are fine, but to be competitive beyond approach, your organization must build propriety models that others cannot duplicate.
As such, the data scientists that you employ on your Big Data strategy team must become more than problem solvers-you need analytic inventors. This takes a special breed of data scientist that can supersede the sophistication of the publicly available analytic techniques, to develop proprietary analyses that nobody knows about. The organizational structure of this team looks more like an R&D function than anything else.
Furthermore, the work that this team is doing—and even the existence of the team— must remain a secret to the outside world. Many components of the organization's strategy should be communicated for broad-based understanding and adoption—this is not one of them if your plan of attack involves causal ambiguity. Remember, the goal is to keep the outside world guessing. Even if they don't know your exact formulae, telling them that you have an R&D team that's specifically focused on developing proprietary analyses is giving off too many clues.
If this is your direction, it's important that you hire data scientists with the intent of retaining them for life. After you have proprietary analyses deployed into your strategy, the last thing you want is a competitor poaching one or more of your key data scientists. This is very difficult to control, especially in today's times when it's typical for talent to change employment every two or three years.
Take adequate measures to ensure their happiness on the job. All of the contemporary human resource strategies apply such as engaging job design, lifestyle considerations, and an outstanding reward system. The most important consideration however, is making sure they have a good relationship with their management team. Most people don't leave a company—they leave a boss.
For your eyes only
In spite of all your efforts to create an enjoyable environment for your key resources, there's no way to guarantee their permanence. Therefore, it's important that you protect your intellectual property in every way you can. Of course, you should work with your legal department to secure intellectual property rights; however, it shouldn't come to legal matters if you institutionalize the confidentiality of your big data capability.
If there's an obvious breach of intellectual confidentiality, your defector should not be able to claim ignorance. Structure your formal and informal systems in such a way that if your proprietary analyses were to make it into the hands of a competitor, it would require a conspicuous violation of both legal and ethical canons. Communicate regularly on the importance of keeping your secrets private, and develop a culture where secrecy is a highly held value.
Finally accept that your secrets may get exposed and take appropriate measures to remain competitive. Keeping secrets in the information age is tough and competitive intelligence may get the best of you. As such, it's important to continually research and develop new analytic techniques. If the old techniques get exposed, the new techniques will keep the competition baffled.
Causal ambiguity isn't the only technique for using big data to your competitive advantage—but it's a good one. The decision to take this path however, requires a long-term commitment to finding the best talent, putting them to work on propriety analyses, protecting your intellectual property, and above all—keeping everything a secret. If your secrets are exposed too early, you may be the victim of an unexpected vanishing act.