Forecasting Doesn't Require a Crystal Ball

I love everything about New Orleans. It is a city that swells with great music while inflating waistlines with fantastic food. I see Trombone Shorty every time he comes to town. I’m trying to put a crawfish boil together for my friends. I usually go down for at least a weekend of the Jazz and Heritage Festival every year. I love jambalaya, both the food and the song. I’m writing this while listening to my “NOLA” Spotify playlist. I wish they would sell Zapp’s Voodoo potato chips where I live. Although I don’t have ready access to the chips, there is a little voodoo we can do up here: Use VMI data to help with forecasts.

Double, double, toil and trouble; Datalliance eases inventory trouble

Having access to VMI data is better than tea leaves or tarot cards. VMI works by looking at what is being sold one step down the supply chain, one step closer to the end user. All products have an end user and the closer the data source is to that user, the better the forecast will be. What makes VMI data especially good for forecasting is that it captures another piece of data. We can take the inventory level desired from an inventory carrier and determine a secondary driver of purchases. Combining these two pieces of information allows a supplier greater insight into what is driving their sales.

There are effectively two reasons sales increase. Either the sell-through has increased or the distributor carrying inventory is putting more onto their shelf. The first reason is great, it means the end user is buying more from the distributor, thus the distributor is buying more from the supplier. Any sustained increase in sell-through will eventually have to hit the bottom line of the supplier who made that product. There can always be spikes, but this type of increase is much more fundamentally sound than changes in inventory levels.

Sales lifts driven by a partner’s changes in inventory levels tend to be more transient. The biggest concern is whether these inventory decisions were active or passive. A customer who has made the conscious decision to increase inventory will increase purchases from the supplier while the buildup of inventory occurs. When the customer reaches their desired level of inventory, which can take months, sales will drop to pre-buildup levels. We have seen a lot of that over the last few years. As the economy improves, distributors and retailers are less concerned with being stuck with inventory, the price of money has dropped making carrying costs less expensive, and more stability offers the chance for differentiation around service levels. The much worse situation is if the customer hasn’t made an active decision to carry more inventory. When someone at the reseller recognizes that there isn’t the demand to support the inventory levels, a sell-off or a series of returns will start. To get an idea of what that looks like, think back to 2008.

Suppliers’ forecasting is typically done with sales data alone. Best case, this is done with customers’ point of sale data but far more commonly it is only with sales from suppliers. In either scenario, how can the demand from end users be differentiated from a change in stock levels? And this is forecasting, when will this change in demand become a change in purchase orders? Incorporating inventory data can help answer both questions.

Datalliance has a report we call “Order Point vs Inventory Value.” It might be my favorite graph we produce. It displays the currency value for both order point and inventory on the same chart. I like it because Order Point is a pretty good proxy for sales. It might even be better than raw sales data because it uses sales data after spike analysis.

The chart clearly shows how sell-through and inventory are moving in relation to each other. The closer the two lines are, the more quickly sales at the distributor will become sales at the supplier. The more important information the chart conveys, however, is determined by the shape the two lines make. If they are moving together, or sell-through is increasing faster than inventory, the underlying cause of any sales change is end customer demand. If, on the other hand, sell-through and inventory are diverging it is the inventory that is driving the sales change. Typically, if the order point is constant or increasing the inventory change is a deliberate decision.

If the sell-through is decreasing it is probably a passive decision and trouble is brewing. Accidentally building up stock usually leads to a sell-off when distributors try to burn off excess inventory. During that time orders to the supplier will drop. If returns are available, large amounts of material may come back as well. Seeing this shape can serve as an early warning sign that a period of weak sales or large returns is on the horizon. This is what we saw in 2007 before the recession.

I'm not Marie Laveau, but this report gives me an at-a-glance insight into a subject that is often shrouded in mystery. It is part of our ongoing quest to give our customers a deeper insight into their supply chains. Think of it as a little bit of supply chain voodoo, no crystal ball needed.

About the Author

David Hall is an Analytics Project Manager at Datalliance. He has been working in supply chain since 2007, spending several years consulting in the hospitality industry focusing on inventory control and supply chain management. Today he is focused on customer engagement and success metrics.To learn more about this report and how it can give you insight into the health of your sales without a crystal ball, call our office 513.791.7272 or email David.

View All Blogs