Subscribe to the Supply Chain Planning Blog

Keep up with the latest trends, research, and insights about supply chain planning, demand forecasting and inventory optimization.

← BLOG

Supply Chain Innovation: Data, Data, Everywhere …. But Where to Begin?

By Jeff Bodenstab21 Oct 2014

Last week we reviewed a Gartner keynote presentation on how supply chains are becoming more data and demand analytics driven. But despite the opportunity, some may be wondering, “Where to begin?” So this week, we’ll look at small first steps companies are making to achieve significant leaps in their demand forecasting.

Let’s start by picturing a typical consumer goods manufacturer looking “downstream” from its warehouse Distribution Centers (DCs) through the retailers’ supply chains to the end consumer. We see hundreds of retailer ship-to locations. Beyond that, thousands of retail points-of-sale (POS) and millions of individual POS scans. Each stage represents a rich source of demand and logistical data, expanding by multiple orders of magnitude.

Each supply chain stage adds useful data

Most companies have lots of this demand data, but haven’t figured out how to use it. So instead, the demand is perceived as noisy and volatile. To improve forecasting, organizations can harness this data already at their fingertips to distinguish and extract the demand signal from the noise.

Here are three suggestions on where to begin.

First, one valuable source of available data is line-orders. Most companies forecast future demand and determine inventory targets by analyzing aggregated demand history. For instance, their forecasts may be based on historical monthly demand quantities by product (SKU) at their own DC (ship-from locations).

There is more information in order-lines than aggregated demand quatities 

Yet almost all companies maintain much more detailed demand histories in the form of individual line-orders. So they already know whether their monthly or weekly demand for 48 cases was generated by one very large order for 48 cases or by 12 smaller orders with an average size of 4. That is, they know line-order frequency and line-order size, two crucially valuable pieces of information to understanding the statistical behavior of their demand. With this understanding, they have the fundamental building blocks for both predicting future demand and determining how much inventory is required to absorb the demand volatility in order to guarantee desired service levels.

Line-order frequency and size are also particularly helpful in understanding volatile forms of demand, such as “lumpy” demand, vertically-integrated or multi-channel demand, or lots of new products or seasonal changeovers.

second source of insight is “Ship-To” data. Many organizations ignore daily “ship-to” orders and use aggregate demand quantities at the DC level to calculate and consume the weekly forecast. But demand data (orders or shipments) at the “Ship-To” or VMI location contains valuable predictive information that can improve the short-term forecast. Customer ordering patterns can be identified (e.g., Wal-Mart places big orders on Thursdays, except during the last week of the month). These patterns improve the forecast consumption logic, compared to empirical “backward-forward” rules.

This information also helps address another problem with aggregated demand streams: One customer’s orders can consume another customer’s forecast. “Ship To” data enables consumption logic that generates better residual forecasts for calculating daily replenishment and deployment.

third source of insight is downstream or channel data, which can improve forecasting by extending the supply chain visibility to the global network to take advantage of daily sell-out data and store-level or retailer DC inventory positions. This downstream data helps to reduce the uncertainty of expected retailer orders to better understand customer behavior and translate it into upstream forecasts. Although POS and retailer data streams can pose hurdles, many solutions are increasingly capable of addressing data collection, harmonization and data management challenges. Costa Coffee, as discussed in a (previous blog) is an example.

Most companies already have lots of data to start building more robust demand forecasts. Small steps to use this data more effectively to improve forecasting can quickly lead to lower optimized service levels and inventory levels.


For an infographic and short whitepaper and on steps companies are making to achieve significant improvements in their demand forecasting, click below.

Subscribe to the Supply Chain Planning Blog

Keep up with the latest trends, research, and insights about supply chain planning, demand forecasting and inventory optimization.

Supply Chain Brief