Home > Articles

  • Print
  • + Share This
This chapter is from the book

1.3 Operationalizing Big Data from Global Perspectives

As the world gets closely connected through digital media and machine-to-machine (MTM) technology with accelerated globalization, the volume of data that is being generated today is astronomical. In 2012 alone, an estimated 2.5 zettabytes of data were generated across the world, and the volume of business data across the world is expected to reach nearly 45 zettabytes by 2020 (A.T. Kearney 2013). At this pace, the world will be inundated with data every day at the rate of an additional 2.5 quintillion bytes of data per day (IBM 2012). This deluge of data or information overload creates unique managerial challenges that today’s business executives rarely faced in the past. What may not be keeping pace with the explosive growth of data volume is the absence of clear business strategies that can handle big data, and the lack of education and training programs for managing big data, not to mention adequate information infrastructure. Another challenge includes a difficulty in sorting, determining, and keeping data that business executives can trust among a sheer volume of raw data. So the fundamental question that we are raising is how to deal with this onslaught of data, while figuring out which business value to be extracted from this big data. In other words, how do we make big data a competitive differentiator for the organization? To answer this fundamental question, we may need to take the following steps:

  1. Identify: Before leveraging big data as a driver of business value, we need to first identify who will be using the data. The identification of target users will allow us to determine which type of data is worthy of consideration for collection and storage. Afterward, we need to identify where the data is coming from, who is creating those data, and where the content lives.
  2. Filter: Depending on the purpose of data usage, we need to determine which data is relevant for analysis and which data should be thrown out. This step will prevent the decision maker from using outdated and irrelevant (misleading) data.
  3. Analyze: After a manageable number of datasets are created, the next step to take is to determine which information should be extracted from the given datasets to solve particular decision problems encountered by the business executive. For instance, if you are interested in finding the optimal routes for package delivery couriers, you need to extract information about the costs/distances associated with each route option and capacity limits of each available courier. The type of information that you are looking for will dictate the choice of various data analysis tools such as descriptive, exploratory, inferential, predictive, causal, and mechanistic analysis. According to Smith (2013), descriptive analysis focuses on the quantitative summary of data features (e.g., mean, median, standard deviation, frequency, cumulative percentage). Exploratory analysis is intended to find previously unknown relationships. Inferential analysis is designed to test theories about the nature of the world in general (or some part of it) based on samples of “subjects” taken from the world (or some part of it). Predicative analysis is intended to make predictions about future events using current facts and historical trends. Causal analysis aims to find out what happens to one variable when another variable is changed. Mechanistic analysis helps us understand how changes in certain variables can lead to transformations in other variables through iterative experiments.
  4. Disseminate: After insightful information is extracted from data analysis, this information should be transmitted to the right person at the right location at the right time. Especially for confidential or proprietary information, information security should be ensured to avoid information breaches.
  5. Update: To avoid a wrong decision stemming from outdated information, we should constantly keep data updated and monitored for veracity.

Following a closed loop of deriving value from big data and then gaining insights into business decision problems, we should not overlook the process of operationalizing big data by putting those insights gleaned from big data into use or practice in a form of repeatedly usable solutions. One way of facilitating that process is to develop visual reporting mechanisms (e.g., summary reports, graphical charts, tables, dashboards) that can be easily shared by the team of decision makers and stakeholders without technical expertise. More important, since intelligent insights obtained from big data analysis may not be shared throughout the organization, we need to first break visible or invisible silos that disconnect some organizational units partially or wholly from the rest of their organization.

  • + Share This
  • 🔖 Save To Your Account