RFID and forecasting: Where's the benefit?

Send to friend

The RFID (radio frequency identification) bandwagon is being driven into town, with retailers and some suppliers claiming that the technology will bring significant benefits in the areas of on-shelf availability, reduced shrinkage, lower inventories and improved promotion management. However, for many suppliers, the case for RFID is less certain, and relies on benefits claimed to result from improved supply chain collaboration, forecasting and demand management. What evidence exists for these claims and where might the benefits be found?

Will RFID improve forecasting and demand management?
The premise underlying the claims that RFID will improve supply chain collaboration and demand management is that a great deal of detailed and accurate information will be made available for each case of product at each location in the supply chain. Since this data is more detailed and more accurate than the information we currently have, then our ability to manage demand will improve and everyone in the supply chain will benefit as a result

The weather forecasting analogy
In weather stations across the country we have instruments measuring the current weather situation. This real-time logging of detailed weather data is very similar to the way in which RFID will make real-time inventory data available to the supply-chain. So, for our premise that more information of a more detailed nature helps forecasting to be more accurate, we would expect that the availability of this real-time weather information would have improved our weather forecasting in recent years.

The reality is that short-term weather forecasting has improved and is now quite reliable. The availability of real-time, high-quality current weather data has undoubtedly contributed towards this, since the weather we will experience in the short term develops directly from the weather that we are experiencing now.  However, as we move out in time, this information has less and less value. For example, the fact that we know that we have had 3 mm of rain in the last hour does not help us to predict a "white Christmas".

Returning to the supply chain situation, the intuitive conclusions for the use of RFID data based on this weather forecasting analogy would be:
 We should expect our short term forecasting and management of demand to improve, since our understanding of the starting point of the forecast has improved
 We should not expect that real-time data will assist longer term forecasting, since it will be superseded by other more relevant information over time.

"RFID data shares the same characteristics as EPOS data in that it is indirect, expensive to process, and is only likely to be available from a small number of retailers in the next five years."

Lessons from EPOS data
Further insights may be gained from the availability and application of EPOS (electronic point-of-sale) data. Despite a few well-publicised early cases, the reality after several years is that the data is used sparingly and this is for a number of practical reasons:
 Suppliers are measured on their service levels in response to the replenishment orders of the retailers. They are NOT measured on their ability to satisfy consumer demand
 EPOS data is "indirect" data-i.e., data that is not used to drive the primary forecasting model (the direct data is usually retail replenishment orders)as such it requires analysis and judgement before being used in a forecast
 EPOS data requires purchase, translation and consolidation before it becomes useful information. It is therefore more expensive than the direct order data
 EPOS data is only available from a limited number of retailers-it, therefore, cannot give a total demand picture across all customers as seen by the supplier.

EPOS data is, therefore, used primarily in short-term situations such as promotions, in which the retailers initial replenishment order quantities give little indication of the true consumer demand, and the risks of waste and service level issues are greatest.


"if RFID becomes one part of a greater scan-based trading scenarioboth EPOS and RFID data will become critical to suppliers as part of their total supply chain solution"

RFID data shares the same characteristics as EPOS data in that it is indirect, expensive to process, and is only likely to be available from a small number of retailers in the next five years. If retailer/supplier partnerships continue to develop at the same slow rate as the last five years, then the benefits of RFID in demand management are likely to be very limited. However, if RFID becomes one part of a greater scan-based trading scenario in which the supplier manages his own stocks in store, then the retail replenishment order will cease to drive supplier performance. In this case both EPOS and RFID data will become critical to suppliers as part of their total supply chain solution and will be fundamental to their ability to serve their consumers via retail outlets.

Richard House is the managing director of FuturMaster Ltd., a leading provider of forecasting and supply chain software solutions for manufacturing and distribution companies. He has over 15 years' experience of supply chain planning and forecasting gained in consultancy, line management and software solutions. He may be contacted at rhouse@futurmaster.com

Comments (0)

Add a Comment

This thread has been closed from taking new comments.