point
Anupam Dasgupta

Senior Manager

Visual Bi

Extinction Of Linear Equations !!!

   Over the last eight years I have been dabbling with various facets of business development, business analytics and corporate planning in IT/ITES. Business problems, whether significant or trivial, more often than not, require a multi-pronged approach. Or let me put it this way, there are hardly any linear equations in the real world!!

Multiple parameters, with varying degrees of significance, accompany most, if not all business issues. Which makes me think of one of my favorite subject areas that never fails to fascinate me: Management Consulting.

This is 2013, well almost 2014, and here we are with analytics firms booming; data-driven business insights are in demand and analytics provides the required rigor and accuracy. On the other hand you have the classical strategy consulting firms that operate through consultants equipped with subject matter expertise in management disciplines and sometimes also with specific industry experience. Let me elucidate further, you could have a consultant creating entry strategy for an European retailer in South Asian markets  – it is quite likely that the consultant has her marketing concepts perfected by virtue of a premier business school background and industry knowledge ingrained due to a stint with a leading retailer.  But then let me get back to where I started.

I am pondering about the perfect approach to problem solving that a management consulting firm might want to adopt. This is more like a proposition that an entrepreneur wanting to start up a management consulting firm would want to figure out!  Statistical analyses without industry knowledge and/or management expertise could create a situation where you have a lot of complex mathematical models and yet the application of the results of the model are misplaced in a business scenario. Again, pure consulting without the statistical rigor of analytics might lead to a situation wherein specific and unique trends in a particular business are overlooked.

This makes me think that a mix of both is the way to go! Once the business problem is defined and the associated questions are figured out, the information required to get to the answers needs to be nailed. Such information should be backed up by sound and comprehensive analytics. Once statistical tools and techniques enable the required number crunching and resultant information, it is on the platter for the management consultant to apply her management expertise and industry specific experience!! The facts, figures and analyses are all laid out and waiting for the management/industry-specific spice to be sprinkled!!

Let me take a hypothetical scenario, simplified for the sake of brevity. John is an e-retailer keen to prioritize his investments between two service lines, books and apparel. Statistical forecasting models predict the same percentage gross margin from both service lines. So should John give identical priority to both? Most probably NO!! He needs to know the specific industry dynamics, possibly a particular gross margin percentage could be spectacular for books and lackluster for apparel!! He also needs to have a sound understanding of the firm’s strategic goals and financial position in order to decide if he wants to focus on the over performer or pull up the laggard?

One can argue that it’s quite possible to configure a mathematical model that will factor in all relevant parameters and provide industry specific recommendations corresponding to different margins. Well yes, I agree, it could be possible! But just that determining the parameters for such a model and setting up the recommendations would require sound inputs from a “conventional” management consultant, who would utilize her management expertise and industry experience to make it happen!!

On a different note, business analytics firms do have domain specialists who try to bring in domain/industry-specific flavor. The capability of such specialists, to scan the customer situation, identify and define the business problems, drill down to the right questions, identify the information needs and thereby define the direction for statistical analyses is critical to be able to serve as a true management consultant. In addition, the specialist should also be able to apply management concepts and industry flavor to the output of data analyses.

In a nutshell, I would theorize that management consulting is driven by 3D’s namely, Data, Domain and Disciplines of management. Ideally, crack data scientists along with true-blue strategy consultants would together constitute the magic potion called management consulting!!

So after all, looks like, it does make sense to have a blend in place instead of going ahead with a one-dimensional approach. So then linear equations are only for an imaginary world, y = mx + iC anyone?!!