The last two weeks I talked about evaluating reporting tools, the different tools criteria categories, and determining the particular important sub topics that are unique to your company. This week we are going to talk about the Big Data itself, and four ways to determine the best areas to cover in your analytical research for maximum return on investment.
- Inventory your current analytic research data. Starting with a base is always the best way to improve a situation rapidly. As I have written in the past, inventory the existing metrics that the business measures. Now, research needs to be done on the elements of the metrics and the context in which they’re used for the criteria, measurements and business usage of the analytical answers. Find out from the business users how they use this data, and what actions they take based on those analytics. Determine whether any of those Big Data analytics can be further extended, enhanced, or exposed to other departments or activities in the business.
- Determine the data domains that have the most improvement potential. Using the Big Data analytics researched in the first phase, determine the best cases for improvement. Do you need more customers, more retention, products, features, locations, or different sizes or types of locations? One example of this type of Big Data analytics is the analysis recently done by several retailers which led them to determine that the size of their stores was too large. They downsized them to handle limited numbers of products per location and sized each more appropriately for the local business activity. Rightsizing and localizing the business for the customer base and their needs can save huge amounts of logistic, property, and employee expenses, turning locations or regions more profitable to the bottom line.
- Improve and aggregate data. Also, using the inventory of the Big Data analytics done in the first phase, determine where some of the many new types of data could enhance your company’s Big Data analytics. Social data, “tweets,” “LIKEs,” sensors, mobile device advertising, underutilized existing digital assets, customer demographics, attitudes, transaction records, time based analytics, RFIDs, GPS, and location-based data is all available for your Big Data analysis.
From sensors in airplanes, farm equipment and loading dock forklifts, location, engine RPMs, and other activities can now be harvested for fine grain, time-based analysis and cross-referenced to location and external temperatures, weather and other work condition factors. Customer sentiment through Big Data analytics of LIKEs or product related tweets can give early notification of product problems, or defects, and help companies fix and respond faster to the marketplace. Even harvesting Google flu type queries are now being used to track infectious disease outbreaks. Determine what data elements you have and determine where any of these new Big Data types could aggregate and enhance your Big Data analytics. - Correlate the data. With your existing elements and enhanced Big Data aggregates, compare, contrast, and analyze the data for patterns, tendencies, mean time to failure, and other business outcomes. Use all the elements, and test all possible subsets to relate the data against your return on investment ideas for your Big Data. Compare and analyze all the existing old elements and the newly acquired “LIKEs” tweets or other data Big Data elements against each other. Analyze element subsets, different element combinations and other element relationships to uncover the keys that will help you discover new product factors, customer tendencies, and other business factors. Sometimes these element combinations and correlations can lead to an additional improvement of sometimes 50 to 1000% or sometimes only 5 to 10% which can mean a profit for your company and your Big Data analytics efforts.
Relating different aspects of your data to each other can be very difficult; incorporating other Big Data sources can even be more difficult with its poor data quality, unstructured elements, and large quantity of unrelated meaningless data. But making the effort to incorporate it can pay big dividends and give your company the extra information and a competitive edge, making the journey of Big Data analysis well worth the effort for a tremendous return on investment.
Dave Beulke is a system strategist, application architect, and performance expert specializing in Big Data, data warehouses, and high performance internet business solutions. He is an IBM Gold Consultant, Information Champion, President of DAMA-NCR, former President of International DB2 User Group, and frequent speaker at national and international conferences. His architectures, designs, and performance tuning techniques help organization better leverage their information assets, saving millions in processing costs.
Leave a Reply