Downstream It’s a Forecast; Upstream It’s a Fact

THE HEADLINE FOR THIS COLUMN may seem like a riddle, but it’s intended as a succinct way to characterize the flow of forecasted order quantities up and down the supermarket supply chain.

The prevalent method has long been to order product in a consolidated way for use downstream – that is, for distribution to stores. However, several retailers have embraced the power of point-of-sale data by originating order quantities from the end of the supply chain – from the stores – and using this information to determine optimized quantities for facilities upstream in the supply chain.

Why This Is Important

Determining shipment quantities and timing depends on an understanding of actual demand and perpetual inventory – with the only true measurement of demand occurring at the cash register. Forecasts generated at the D.C., warehouse or manufacturing levels are by definition estimates of a possible downstream reality. I think of it like fishing…before sonar.

If you’ve ever fished without sonar, it may have gone something like this: Either you or someone you knew had some history with the place you were fishing. You took a guess at where to fish not knowing if there would be fish; even if there happened to be fish in that spot on that day you certainly had no idea how many or what size. You chose the best lure based on what you’ve been told or what you personally knew about the area. You dropped your lure in and hoped for fish not having any idea what was swimming around your lure – if anything.

This process is analogous to today’s prevalent product procurement practices for tier 1 consumer packaged goods suppliers (wholesale or retail owned distributors). Most are still determining order quantities based upon history of what their customers have pulled from their warehouse (like the history of the fish that were pulled out of a certain fishing-hole). In all likelihood, these same companies have been told that their systems use, “a multiple forecasting” technique or approach.

Clean out Your Tackle Box

In all likelihood, your product replenishment system is using several trend-based forecasting techniques combined with an error method to determine order quantities. This is quite like deciding where to fish based on historical trends.

Today’s trend-based ordering processes regularly fall short regarding promotional, display, weather-related, event-related and other aberrations nor do they handle seasonal trend ordering very well. When these factors arise, the buyer typically has no idea how much product is already in the supply chain nor whether the pull-based numbers accurately reflect seasonal, event or ad-related performance at the cash register.

In other words, we buy product for entities that are upstream in the retail supply chain by estimating the quantities needed for the end consumer based on historical data of pull-based demand at wholesale. Downstream it’s a Forecast (and often just a guess).

Fish Where the Fish Are

Modern sonar has advanced to a level today where it can give you a very accurate estimate of fish quantity, depth, size, and even the path that the fish are traveling. By building a Demand Based Supply Chain (DBSC), we can execute the analogous equivalent to this in product procurement…and even tell you which promotional “lure” to use!

A DBSC utilizes advanced statistical, causal-based point-of-sale forecasting along with supply chain perpetual inventory figures to drive purchasing decisions at every entity in the supply chain. As long as perpetual inventory is accurate and forecasting is advanced, product needs can be determined with enough lead-time to move the product through the supply chain – Upstream it’s a Fact.

The results of building a DBSC are phenomenal:

• Warehouses become true distribution centers – averaging between one and two days throughput on most product.
• Sales increase due to improved in-stock positions at the distribution center and retail
• Outbound truckloads depart 100% full 100% of the time.
• Warehouse automation is enabled due to eradication of inventory.
• The “bullwhip” effect disappears completely from the supply chain.
• Companies can realize net-negative inventory by selling product before they pay for it.
• Merchandising dollars are spent on merchandising product and not cleaning up problem inventory.
• New products reach the retail shelf in record time.

Oh, and about the lure…While you’re at it, the information can harnessed for data-mining and merchandising planning.

Eric J. Smith (eric@supplychainopt.com) is Proprietor of Supply Chain Optimization, LLC (SCO), an Iowa-based consultancy focused on store-focused supply chain management, and an Itasca Retail consulting partner. He is also an instructor of Supply Chain Management at Iowa State University. © Copyright 2016, Eric J. Smith. Republished by permission from his blog: SupplyChainOpt.com.