Different Approaches to Analytics Can Transform Logistics
By: Logistics Viewpoints April 11, 2014
Three stages are commonly used to categorize an organization's maturity in their use of business intelligence and analytics technologies: Descriptive, or what happened in the past? Predictive, or what will (probably) happen in the future? Prescriptive, or what should we do to change the future?
Descriptive analytics typically means good old-fashioned business intelligence (BI) – reports and dashboards. But, there is a newish technology in the Descriptive category that one may argue is worthy of a category in its own right. That technology is visual data discovery. This approach has a rapidly growing fan base for many reasons, but one stands out: It increases the probability that business managers will find the information they need in time to influence their decisions.
Traditional BI reports and dashboards can be very effective at helping managers to find the answers to predictable or anticipated questions. For example, a simple text report can provide a list of all items that a distribution manager needs to put on a particular truck. Likewise, a gauge on a dashboard can provide an easy-to-assimilate view of on-time delivery performance, showing how the current value compares to acceptable performance limits. In both cases, the BI solution is providing answers to expected and well-defined questions – questions that have existed since the dawn of logistics management. It makes perfect sense for the algorithms needed to answers those questions to be defined and incorporated into a BI solution. Typically that work – defining and building a BI solution – is undertaken primarily by IT staff with technical skills. And that’s precisely the problem. The conventional business intelligence solutions often aren’t flexible enough to help managers answer unexpected questions – and IT staff isn’t able to make changes to those solutions before managers need to make their decisions.