Tuesday, October 1, 2013
Big data or big confusion?
Ever since we entered the world of “trendy” in SCM class –from lean to reshoring and now big data– I’ve taken an interest in challenging all these hip new words and question everything that they entail while assessing their effectiveness. This week is no different; I want to play devil’s advocate against big data and get the reader to look at this topic through a different perspective.
First, we need to establish a definition for this concept; why is this data big? Big data refers to the software engineering term that describes sets of data that grow so large that they are difficult to process using regular database management tools . As companies realized that they could use data collected from external stakeholders –consumers, suppliers, etc.– they started developing other means to process and capture these enlarged numbers of data to factor in their decision making processes.
These stories of success have caused other companies to pursue similar actions, but they have not been as successful in doing so. Why is that? The data is out there for everyone who needs it to capture; how can Amazon be gaining competitive advantage while others are either losing or not generating as much ROI as expected? There are three main reasons that I narrowed this down to, based on my research:
It’s not magic
As Teradata’s Stephen Brobst cleverly stated, “New technologies are often perceived as silver bullets that will solve all problems” . This topic has been heavily discussed in the past and Nicholas Carr caused quite a stir with his “IT Does not Matter” paper; he has a point though, one cannot continue to do things the same way, invest in a million-dollar technology that everyone is using to improve the business, and just sit and wait for it to solve itself. There must be a good reasoning behind the investment and an awareness that these are investments that might not yield returns until after a series of changes in different areas (culture, processes, technical expertise, to name a few). The entire change cycle could take years before positive results are shown. Furthermore, the investment is not to be done for the sake of having the latest and most expensive technology; the right tools are needed to handle and make the most out of what big data brings.
You need the right people
Part of these tools includes people. Technologies such as Hadoop, MapReduce, and NoSQL that can process big data are extremely difficult when compared to the traditional database-oriented tools. The gurus who know these technologies are scarce, and those who are receiving formal training in schools don’t know enough, so there is a demand from the companies that is not being compensated by what the market has to offer. There are even new job roles that companies have created, such as data scientists, who focus on studying and reading this data to interpret what exactly it is telling them that can be of value. What good does it make to have all this data if you don’t have the right tools to read it or people to analyze it?
You need the right processes
The other part of the before mentioned tools is processes. The decision to invest should have been carefully reviewed and questions such as “what workloads in our processes is this technology most efficient for?” should have been answered. The entire environment should be designed so that the selected technologies work well together with the existing resources. It is also important to assess whether
Not to say though, that big data is bad idea, but I believe that companies do not understand the investments necessary for it to be successful. My recommendation for companies would be to practice smart investments, they need to go beyond technology and consider people with the right skill sets along with the processes. They should think about the value that this can bring to their business when they align technology with the strategy instead of focusing solely on technology. On that note, I ask the reader: Is big data here to stay?
by Elisa Taymes