In the last of my “Mr Analyst” series (see previous 2 articles), I turn my attention towards the issue of centralisation and offshoring of business analytical activities.
With the unevenness of talent and man-power cost across Asia, many organisations have set up offshore “centres of excellence”, hoping to exploit these advantages to scale and accelerate their analytical activities. Believing that they can provide consistent and unified analytical support to the operating countries across the region, they have invested heavily in these centralised set-ups. Some have dismantled or scaled back on their in-country analytical organisations to compensate. The outcomes have been less than wonderful. On the contrary.
In many cases, these offshoring and centralisation efforts have reduced the analytical capacity of their organisations
At the heart of this anomaly is the confusion about the nature of problems and problem-solving. In a broad sense, all business problems can be classified into one of two categories; they are either uncertain or equivocal. A problem is classified as uncertain if the acquisition of incremental data reduces its uncertainty. Consider the example of credit scoring - the more your know about someone, the more you would be able to predict his capacity and willingness to repay a loan. On the other hand, a problem is classified as equivocal if it has multiple and oftentimes conflicting interpretations, and the acquisition of incremental data does not provide clarity. Consider the case where an organisation seeks to respond to a new competitor - the C-suite likely has differing views on the matter.
The reality is that most business problems are equivocal in nature. Because organisations are run by people, and people view problems through different lenses and perspectives that have been shaped by their experiences. Equivocality needs to be actively reduced through convergence of interpretations before the uncertainty challenge can be addressed. Most business analytic functions are poor at recognising equivocality (see my previous article on data scientists). They jump right into the task of solving for uncertainty, committing to an interpretation that may ultimately be irrelevant. Centralisation and offshoring compounds this problem by restricting face-to-face dialogue with and access to the line-of-business; destroying value rather than sustaining it.
Many organisation staff their centres of excellence with data scientists and statisticians. They confuse problem-solving with predictive analytics since so much (digital) ink has been spilled on its benefits. In truth, predictive analytics solve uncertain problems; where the problem is well-defined and convergence of interpretation has been achieved. In fact, predictive analytics is really about optimisation - removing the wastage of poor targeting or classification. The much-touted “Big Data solutions” are almost entirely about solving uncertain problems, which explains the limited ROI in many industries.
Solving equivocal problems requires observation, discussion, and domain experience. Often it requires face-to-face interactions as ideas and concepts are difficult to convey through emails and telephone conversations. Foremost, it requires recognition by the business analysts that multiple interpretations and solutions are possible, rather than a single optimal one. Solving equivocal problems requires finesse with hypotheses development and contextual data mining, all of which are in short supply when you centralise and offshore.
So if you are thinking about making that leap to centralise and offshore, consider the current activities that your business analytic teams are engaged in. If they are already engaged in solving complex, ambiguous, or even strategic problems, you will be best served by having them remain in-country or embedded with the businesses.
Data without interpretation means nothing
Creating the necessary environment to strengthen and support the interpretive skills of your business analytic teams will reap far more dividend for your organisation. I call this "the Power to Solve".