Wednesday, September 3, 2014

Demand Sensing – A modern forecasting technique

A blog post by Preeti Havaldar

“The most reliable way to forecast the future is to try to understand the present”
– John Naisbitt (author of Megatrends, New York Times bestseller)

Though traditional forecasting methods such as time-series forecasting are good indicators of future sales, in practice, only half the consumer packaged goods items sold in North America have the required two years of historical data necessary for proper statistical analysis that accurately accounts for seasonality. This leads to forecast errors and they particularly become ineffective during volatile markets. For example: in events such as a hurricane, the sales of essential commodities like water can spike up unreasonably. To understand this unexpected rise in demand and to deal with it, historical data is not a good indicator to be considered.

Demand sensing is a next generation forecasting method that uses quantitative techniques combined with real-time market information to accurately forecast demand. This intelligent forecasting technique not only helps reduce forecast error by 30% to 40% but also helps mitigate effects of market volatility by facilitating a demand-drive supply chain strategy.

Demand sensing helps cope up with unanticipated situations by taking into account real-time data including orders, shipments and other supply chain data. According to Greg Schroeder, Kimberly-Clark’s senior manager for the company’s supply chain centre of excellence, when a hurricane struck northeast America in 2012, Kimberly-Clark could not forecast the consequent demand accurately. However, now with the help of demand sensing technology, they can analyze daily sales and orders to understand customer needs and hence forecast demand in a better fashion.

As opposed to the traditional time-series forecasting methods that rely heavily on historical data, the demand sensing technique typically uses data analytics to analyze historical data, near real time point-of-sale (checkout) data, inventory levels, daily orders and shipments to accurately predict demand which further leads to efficient inventory management. The diagram below shows the fundamental input-output flow for the demand sensing process:

Profitable growth in volatile markets requires an agile demand-driven supply chain that can quickly react to changing market conditions. According to Nils Mueller, Global IDF Initiative Manager of Procter & Gamble, "In the current economic climate - with ever increasing volatility in demand - it is crucial to extend the supply chain visibility as close to our customers as possible. Speed today is of the essence. Only those companies that can respond quickly to customer and consumer needs will stay ahead.”
Demand sensing platforms are being incorporated by some of the world’s best-known companies such as Shell, Procter & Gamble, Unilever, Kimberly-Clark, Kellogg and others to enhance the demand forecast experience. However, the question that arises is - Is demand sensing just another buzzword or can it bring a revolutionary change to traditional supply chain management?


No comments:

Post a Comment

Note: Only a member of this blog may post a comment.