Your responses to our survey show that most of you (70%) are aware that EDI data can be useful in the creation of business analytics, but only 40% are working with an EDI software or services provider that offers any kind of analytics based on the data they are managing for you. However, about 50% of the respondents indicate they are using analytics of some kind. Your commentary made it clear that not all of the analytics being used are derived from EDI data, and in fact, some of you (a very small number) believe that EDI may not be particularly useful in the production of business analytics.
On a broader scale, only 10% thought that analytics were not at all important, while 55% indicated analytics were very important. I think that anyone involved in making decisions about their business, and basing those decisions on anything beyond their 'gut feelings' is likely to put some faith in what they would consider analytics... that is, information that is derived from reviewing some kind of data.
But... there is a big difference between what we've come to think of a 'slice and dice' analysis and true business analytics. Keith Collins, Manager of Business Analytics for SAS says that "The term 'business analytics' has become cheapened over time, as different organizations have tried to tie their products and methods to the term." He says that true analytics go beyond the 'what happened' and should be used to help organizations know what and how to cross-sell and up-sell. He points to the field of predictive analytics that discovers the probabilities of a particular customer buying a particular product.
Without diving too far into the process, Collins says there are three steps to effective analytics:
Forecasting uses historical data to understand what has happened, while predicting applies the understanding of the forecast to specific customers and segments of customers. This goes far beyond the typical "we sold X number of product #123 last year, and should sell Y this year," and looks at the customer, location, demographics, and other factors that segment the sales.
Optimization, says Collins, "Analyzes the customer acquisition and relationship based on the customer's lifetime value, and allows the development of specific campaigns to fit the needs and patterns predicted for a specific customer, or customer group. This allows for multiple campaigns to be developed to reach each customer in an optimal way."
Can your EDI data be used for this? More importantly, should your EDI data be used for this? Does EDI transactional data provide enough information by itself, to allow this kind of analysis? What other data needs to be available and included with EDI data to make this kind of analysis even feasible?
We'll look at some of the methods and services available and in development next time around. If you know of, or have experience with managing this kind of advanced analytics, let me know. According to experts like SAS' Collins, effective deployment of advanced analytics will increasingly be the differentiator in competitive situations.