Shari wishes to thank Sanjay Mathur and Michael Bechtel for their contributions to this month's column.
In August, this column discussed emerging silent commerce technologies cheap radio frequency identification (RFID) tags and sensors in conjunction with wireless technologies as a means of gathering data to improve inventory control and supply chain efficiencies, and reduce labor costs. According to Peter Abell, research director of global retail at Boston-based AMR Research, RFID may offer labor savings of 20 percent in the warehouse, reduce supply chain inventory by 25 percent and increase sales by three to four percent ("The Path to RFID," Susan Reda, Stores Magazine, June 2003).
How can organizations turn new RFID information into actionable, 360-degree business insight? Researchers at Accenture Technology Labs believe that several trends are converging to help companies act more quickly and competitively.
- Scale: With the ability to track individual objects in real time, silent commerce offers an unparalleled level of data granularity. Access to more granular data helps to improve and refine current analysis techniques, while storage technology improvements make the retention of this data cost effective.
- Processing Power: Moore's law regarding the exponential growth of processing power means that insight applications can be processed at an order of magnitude faster than they were five years ago. Companies can ask questions they could not previously ask and receive answers in real time.
- Analysis: Improvements in machine learning enable more effective data analysis (preventing information overload by focusing only on data relevant to decision making) and decreasing software costs put enhanced analytic capabilities in the hands of the average worker.
Analyzing and Acting
Collecting data in a warehouse without analyzing it is akin to hording money without investing it. It's not how much you have; it's what you do with it. The types of supply chain analyses that can be performed using silent commerce data are broadly classified in two categories:
- Descriptive analyses that seek to understand the information we have, and
- Predictive analyses that attempt to predict information we do not have.
Descriptive analyses find facts already in the data. These analyses run the gamut from simple totals and averages to complex association rules and clustering algorithms. Before conducting an analysis, it is critical to know the question being asked of the data, as well as the action that will be taken when the answer is found. For example, a supermarket warehouse uses RFID to track forklift movement through the warehouse. It uses descriptive analysis to analyze its capital equipment spending and learn the average daily percentage that those forklifts are active. If the analysis reveals that the forklifts are stationary for a suboptimal daily percentage, a decision can be made to operate fewer lifts.
Because each material pallet in the warehouse is also tagged, its presence can be detected by lifts outfitted with RFID readers. To further improve warehouse efficiency, descriptive analyses might determine that shampoo and conditioner are high-velocity products (present in a significant percentage of overall loads carried) and are highly correlated (likely to be moved in the same load). Based on these analyses, the shampoo is moved closer to the dock door and conditioner is moved next to the shampoo to minimize the length and frequency of trips.
Predictive analyses leverage existing data to provide suggestions and make predictions. Models guess unknown variables by leveraging trends and patterns that are detected in known data. Silent commerce data can improve functions such as inventory and demand forecasting, where predictive modeling is already used, and can bring predictive analytics to real-time, tactical use. For example, given the visibility provided by RFID, a retailer could access a real-time inventory location long before the inventory arrives. A predictive model could use that data to suggest the probable date and time of arrival for individual items. As a result, the retailer can minimize costly safety-stocks and better prevent potential stock-outs.
Additionally, RFID sensors on retail racks could provide a detailed understanding of items that are purchased, handled and put back, or not handled at all. A predictive model could use this data to suggest more profitable locations or prices for the items. In this way, the retailer becomes more aware of both supply and demand drivers, and can take action accordingly.
Silent Commerce, Insight and the Supply Chain
When sensors and tags are used throughout the supply chain, businesses can access the kind of granular data traditionally associated with customer relationship management (CRM). Just as enterprises know the individual details (name, age) and aggregate attributes (segment, life cycle) of their customer base, silent commerce enables enterprises to track the micro and macro attributes and behaviors of the physical objects that comprise their business. Unlike CRM, however, supply chain analysts do not (yet) have the option of purchasing third-party "demographic information" for their widgets. The data that can be analyzed is limited to the data that we choose to collect. It is for this very reason that Accenture encourages companies to start tagging now so they can eventually ask the questions that offer actionable insight and provide a competitive edge.
Next month, we will take 360-degree insight down the financial path to discuss the nuggets of information that are available in your financial systems.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access