Big Data and Me: The Unlikely Duo

I came across a blog this week that summarized some of the main points in Waller and Fawcett's on big data and supply chain systems. One of the blog's notes include Waller and Fawcett's notion that quantitative/technical skills are not the "end all be all" of data analytics in business applications (Ferguson 2013). An understanding of  the underlying business environment behind data is also a requisite skill for powerfully using data to better a business or supply chain. 

As Carnegie Mellon students, we've all come across classes/situations that address extremely quantitative or technical approaches to business problems. For the past two weeks, I have been learning about multi objective goal programming. In short, this type of mathematical modeling is used to achieve solutions to problems with competing/contradictory objectives. These types of problems do not always have a clear question that is asked nor will the best solution always stay the best solution. Priorities change, budgets change, and consumer demand changes, which all require agile, creative approaches to analyzing data. In a blog entitled, "Predictive analytics-art or science?," the author, Rachel Clinton, emphasizes the predictability component of big data and notes that "Some basic analysis can be done using this data by means of canned queries, pivot tables, summarized reports, and so on— but that’s basic ‘how much, how many, how often’ analysis – it’s not predictive analytics" (Clinton 2013). Spot on, Rachel.  As someone with very little technical background, I find myself able to contribute to complicated MOLP problems by knowing the setting in which the problem takes place, understanding how the variables interact with each other from a qualitative standpoint, and using business acumen to infer what the writer of the problem--in this case the author of the textbook--really wants. But without knowing the ins and outs of Excel, Oracle, or any other database/analysis platform, how can I conceivably contribute?

My Management science professor has often cited example where says that his success as a mathematical modeler has come from creating models that decision-makers can then use to feel like they have come up with the right answer. In fact, the whole point of the course is to become an effective "end user" modeler (Caulkins, Management Science I Syllabus). And nowadays, there are many type software out there that allows decision makers to solve network optimization, MOLP, and other types of problems without necessarily building the algorithms themselves. One of these is just the Solver add in to Excel that Public Policy majors have become familiar--some may say too familiar--with. Sometimes, the person on the end user is the last link to making the right decision in stream of complex business problems. 

I really loved Waller and Fawcett's article because it promotes that idea that varying skills sets are needed in a complex process like making a decision about a supply chain. Waller and Fawcett even notes that for some SCM applications, softer skills like budgeting and evaluating opportunity costs can be more useful than the more theoretical/technical/quantitative skills in the same discipline (Waller and Fawcett 2013). This brings to light an important question, however. How can people in such widely varying disciplines work together while maintaining effective and productive communication lines in a supply chain?

Sources:

Clinton, Rachel. Predictive analytics – art or science? Retrieved from http://www.sv-europe.com/predictive-analytics-art-or-science/.

Waller and Fawcett. Data Science, Predictive Analytics, and Big Data: A Revolution That Will Change Supply Chain Design and Management. Retrieved from file:///C:/Users/arie/Downloads/SSRN-id2279482.pdf.

Ferguson, Renee. "Are Predictive Analytics Tranforming Your Supply Chain?" Retrieved from http://sloanreview.mit.edu/article/are-predictive-analytics-transforming-your-supply-chain/.