By using a technological crystal ball known as demand forecasting software, jan/san distributors are gaining critical insight about their supply chain and how their products may move in the future.

Although a demand forecasting solution has become a key component to a distributor's analytical system, it still must be coupled with good old-fashion internal communications between sales staff, operations managers and finance directors. It also does not replace the need for client and supplier collaboration.

"Accurate forecasting allows us to keep the fine balance between keeping costs low by preventing overbuying while maintaining customer service levels by maintaining sufficient supply to meet customers sales," says Jonathan Soon, vice president of operations at Royal Corp., Santa Fe Springs, Calif. "In like manner, our sales team is constantly looking at customer trends and work with their forecasting to remain ahead of any deviation that might occur."

Currently, there are dozens of forecasting solutions on the market. They range from complimentary solutions that attach to enterprise resource planning (ERP) software to stand alone software suites. Demand forecasting software allows distributors to foresee the future in terms of when certain products will be in demand and when other products will be idle. As a result, implementing demand forecasting software into an operation will allow a distributor to determine where and when to apply financial and human resources into the supply chain.

"The best part of any forecasting software or system is that it will allow you to look at trends," says Jim Smith, executive vice president of Indianapolis-based HP Products. "The hardest thing for anyone to understand is the seasonality of goods. Almost all business works on a bell curve. Nobody stays busy all year. It just doesn't happen."

Finding The Fast And Slow Movers

Some demand forecasting systems gives the user the ability to give products weighting depending on certain product characteristics, such as whether they are seasonal items or whether they are fast or slow movers.

"You don't want to go in to a period of time and not have the business on something and be sitting on $100,000 worth of inventory that is not going to be purchased for another seven months. Today everything is about turn and earn," Smith says. "You have got to be able to turn inventory and you've got to be able to earn money against that. The faster you can turn it at a high gross margin increases your turn/earn ratio."

Demand forecasting software requires the distributor to enter data into modules regarding sales and product movement over time. Distributors also will enter data regarding product ordering as it happens. As a result, the software learns product flow from manufacturer to client as how it impacts the distributor.

"It takes into a rolling account of what is happening with the product," says Jerry Garbett, general manager at Arkansas Bag & Equipment in Little Rock, Ark. "It doesn't look at last year's numbers, it looks at the last 12 months. So it's a rolling average."

Demand forecasting software also, in general, will let the user specify whether products are fast movers ("A" items) or slow movers ("B" and "C" items). Garbett explains that forecasting software has a tendency to suggest keeping slower moving items overstocked for fear of back orders. This forces distributors, in some cases, to buy product before it is truly needed.

"It is trying to have enough in stock so that when an order does come through it is there," Garbett says. "Because what the software is trying to eliminate is us trying to go out there and build an order to fill a 'B' item. Anybody can buy 'A' items, but it's having a mix of the slower moving items is where most people miss the boat."

Soon says one of the features that his company uses is the software's built-in scorecard. According to Soon, his software tracks the company's success, allowing officials within his firm to see how well the system anticipates an item's demand and allows them to make any necessary adjustments.

"The system maintains these forecasting formulas in an accessible table that we can edit," Soon says. "This enables us to custom the formulas for our particular business."

Royal Corp.'s demand forecasting system puts items in six characterization categories depending on their sales history and whether they are a trend item, seasonal item or erratic selling item. The software then uses a series of formulas depending on these categories to continually refine its forecasting. The more historical data regarding a particular item and how well the software does at predicting the item's demand, the more accurate forecasting data it will produce.

The Bullwhip Effect

One challenge that forecasting software users are facing is that data needs to be entered in the system in a timely fashion throughout the enterprise. Many software solutions also require a review of the forecasting formulas on a regular basis and operators who know how to tweak the formulas when needed.

"As in all technological systems, this is supposed to augment the work of a savvy and dedicated purchasing department and not replace it," Soon says. "Communication between sales and purchasing remains paramount in getting the optimum inventory management results."

Steve Epner, founder of Brown Smith Wallace Consulting Group in St. Louis, describes the bullwhip effect as being the human factor impacting demand forecasting, especially when errors happen. For instance, when a back order occurs, supply chain managers who are not careful will view the original demand separately than the order to fill the back order. This will create a false increase in demand leading to a surplus on the shelves.

"The bullwhip effect says that if you are not careful in doing your forecasting, minor inconsistencies in the supply can have an outlandish effect on how much inventory is wasted in the channel or that is not necessary," Epner says. "Then they have too much inventory and it can take years for excess inventory to be used up."

The human element, according to Epner, can also stifle the process by creating a false reality based on inflated expectations. For example, if salespeople hope for higher sales, or if customers indicate a higher need for cleaning supplies because they hope to be busy, then the demand forecast spikes.

"At each step along the way, exaggeration is getting worse and worse and nobody knows what reality is," Epner says. "We have to separate reality from hopes and dreams."

To offset this factor, distributors should rely on forecasting data that is produced by the software, closer to present time, instead of far out in the future, according to Epner. Distributors should also trust the system they have in place and what it is telling them in terms of analytics.


Sharing data across different enterprises up and down the supply chain is one of the major features distributors can take advantage of with forecasting software. According to Epner, distributors should realize they are between the customer and the manufacturer and that these three groups could take advantage of built-in collaboration tools that allow communication between each supply chain player.

The advantage to using collaboration tools within demand forecasting software is it gives distributors, suppliers and customers a clearer picture of the potential flow of products from all angles, not just their own. This can help distributors and suppliers plan for fluctuations in demand and supply amounts and help them plan resources.

In a broad sense, collaboration functionality is a natural extension to demand forecasting software. According to experts and users, the sharing of data throughout the supply chain is the essence of demand forecasting and using these technological crystal balls.

Brendan O'Brien is a freelance writer based in Greenfield, Wis.