SCDigest Editorial Staff

SCDigest Says:

The reality is that even more consumer goods companies, recent sales data is not well used to drive short term planning processes. What Do You Say? Click Here to Send Us Your Comments Click Here to See Reader Feedback

A critical component needed to develop a more “pull-based” supply chain is to be largely driven by actual customer demand signals. Ideally, this should start at “point of sale,” which literally means POS data for retailers and consumer goods manufacturers, but for other sectors, simply means end-consumption data however those transactions occur.

For example, in the medical supplies market, it is currently difficult to impossible to actually receive data from health care facilities about product consumption. Manufacturers in those markets generally can hope, at best, to get shipment data from distributors as a means to gauge actual demand.

Lately, a new term called “demand sensing” has entered the supply chain lexicon. What is demand sensing?

According to Lora Cecere of AMR Research, demand sensing involves the translation of downstream data, with minimal latency, to understand what’s being sold (product attributes), who’s buying the product (customer attributes), and the effect of demand-shaping programs to increase revenue.

“Companies good at demand sensing have minimal latency when sensing actual customer demand and translating it into the supply chain response,” Cecere wrote in a recent AMR research note.

The reality is that even in most consumer goods companies, recent sales data is not well used to drive short-term planning processes. That is starting to change. Procter & Gamble, for example, found substantial reductions in out-of-stocks and inventory levels during a pilot program for new demand-sensing technologies, which lead to a planned roll-out across the company. Based on this demand sensing intelligence, P&G says it may actually adjust the next day’s production schedule 2-3 times to get it more in synch with actual customer pull.

In the consumer goods industry, “With traditional supply chain processes, it takes 7 to 14 days for shelf takeaway to translate to demand in the form of a retail order to a supplier,” Cecere says. “There is lag time accompanied by the bullwhip effect. The more demand shaping — that is, price changes, trade promotion, new product introduction, and sales incentives — the more acute the problem. For example, a 30% increase in sales during a promotion may translate into an order for 30% more product the following week. However, there could be (and often are) massive out of stocks during the promotion. Once the promotion is over, the replenishment order is usually overstating requirements.”

(Supply Chain Trends and Issues Article - Continued Below)

Cecere cites three ways companies especially, but not exclusively, in the consumer goods area, can improve their demand sensing capabilities:

Greater use of vendor-managed inventory (VMI): While strong levels of partnership and collaboration are needed to make VMI work well, the benefits of VMI to both sides of the equation can be substantial. AMR found that the greatest benefits happen when this account-specific demand signal is synchronized weekly to the corporate demand plan.

Use of downstream data by customer account teams: It will take time to well integrate short-term demand signals into back-end supply chain systems and processes. In the meantime, customer account teams, such as those found in the consumer goods industry, can provide benefits for the company by incorporating downstream sales data in their joint planning with retailers.

Daily recalculation of forecasts and safety-stock levels at distribution centers:  This signifies movement from rule-based consumption in traditional distribution requirements planning (DRP) to daily modeling. Cecere says this is the option “with the fastest ROI that most effectively closes the gap on demand sensing.”

In the end, Cecere says that the most successful approach is use of a combination of methods.

“The most successful companies feature those that have experienced high demand variability, have more than 30 VMI relationships (with 50% now based on POS data), and focused on improving the demand signal through the use of statistics to replace traditional, rules-based consumption logic,” she says.

Look for more on demand sensing from SCDigest soon.

SCDigest is Twittering!

Follow us now at https://twitter.com/scdigest