By Bill Bittner, President, BWH Consulting
Many of the innovations at this year's NRF are difficult to understand. These "invisible" innovations include applications ranging from forecasting for merchandise planning and workforce management to security applications for protecting your wireless network from hackers.
As far as the forecasting applications go, the vendors include the big and the small. Among the "statistical giants" are SAS, long known by technology users for the premiere data sampling and reporting tool for tuning IT infrastructure capabilities. They acquired MarketMax a few years ago and are now supporting retail businesses directly. Among the small, are companies like Island Pacific who serve the specialty retail channel.
Utilizing the forecasts involves converting forecast requirements into purchases and deliveries for merchandise and into tasks and employee schedules for labor management. The conversion process can be even more challenging than the forecast and the results need to be continually tuned against the actual outcomes.
Security applications are a whole different issue. We have heard the horror stories and probably all live with the concern that someday our CEO will be reading his own headline in the retail loss prevention press. On the one hand, we don't want to put our information "behind bars" where accessing it is so difficult that employees prefer to make guesses on their own rather than use the systems. Oh the other hand, we have an obligation to protect the corporate assets that are on the network and not allow just anyone access. This is compounded when you consider allowing others, such as DSD vendors, to use the network to improve their store service.
Moderator's Comment: The question with all these "invisible" innovations is - how do you know what is best? Which new forecasting algorithm will give your organization the best results? How much security is enough?
I think these particular examples are two extremes of the types of invisible innovations that must be evaluated by retailers. In the case of the forecasting algorithms, they are all probably good enough. This is the perfect example of where an 80 percent solution utilized 100 percent of the time is better than a 100 percent solution that does not get utilized because it is too complicated or poorly understood. Forecasting systems have to be easily maintained and intuitively understood by the people using them in order to be effective. Although a "standard test" that would benchmark the actual recommendations is possible, I would be very surprised if any of the systems were so far superior that they could make up for poor usability with better analytics.
The security applications are a whole different situation. You can't afford to be wrong in these decisions. The best answer is to have layers of security that prevent full exposure of corporate assets by anyone at store level. Remote users must be validated and their data exchanges encrypted. This all requires the review of an expert and cannot be merely an 80 percent solution. Store managers lose jobs over inventory losses; technology managers lose jobs over security breeches. - Bill Bittner - Moderator