Chemical Manufacturer, Bedoukian Research, Uses SmartForecasts' Intermittent Demand Forecasting to Set Accurate Inventory Levels

Bedoukian Research, Inc., located in Danbury, Connecticut, is a specialized process manufacturer producing high quality ingredients used in flavors and fragrances for customers worldwide. Bedoukian makes 350 specialty chemicals, in small batches, from 2000 raw materials. In addition, the company manufactures fifty insect pheromones used to attract and control insect pests, and resells 100 other products. Two-thirds of Bedoukian's products are irregular sellers that are hard to forecast.

Bedoukian Research bought its first copy of SmartForecasts in 1987 and used it to forecast products with high volume demand. However, it wasn't until Smart Software made available its intermittent demand forecasting capabilities that Bedoukian was able to achieve the level of forecasting accuracy it wanted in predicting its hard-to-forecast products.

The Challenge
Bedoukian Research puts a high value on customer retention. According to its president, Bob Bedoukian, whose father founded the company in 1972, “It's important to our customers that we have good, quick customer service. We need to have the right things in stock when our customers need them.”

While making sure that its products are in stock is a high priority, Bedoukian Research also tries to keep its inventory as lean as possible. Keeping a balance between excellent customer satisfaction and minimum required stocking levels is critical to the way Bedoukian wants to do business. It is also consistent with the service its customers have come to expect.

“Before we obtained the intermittent demand forecasting feature, the forecast results for our slow-moving items were unrealistically flat, not dependable at all” says Leona Eggleston, IT manager at Bedoukian, who also spearheads the company's forecasting process. “For these items, we had to set stocking levels by hand using primarily judgmental inputs.”

Bedoukian has been using SmartForecasts for close to 20 years and is very satisfied with its general forecasting capabilities. However, to achieve its high customer service goals, the company needed a solution that would provide it with accurate stocking level estimates for all of its products, including those with intermittent, or “slow-moving”, demand.

The Solution
In 2001, Bedoukian Research became an early user of the intermittent demand forecasting technology in SmartForecasts Enterprise. This version of the system now enables them to accurately forecast the two-thirds of their products that are not regular sellers, as well as to link the forecasting process directly to company databases. SmartForecasts originally interfaced to Bedoukian's Informix host database. More recently, Bedoukian installed Chameleon, a specialized ERP planning system designed for the chemical industry. SmartForecasts now receives demand history data from Chameleon's SQL Server database and automatically feeds forecast results back into the database.

Each month, SmartForecasts forecasts the demand for every finished goods item four months into the future. In the past, Bedoukian didn't forecast raw material requirements because it was too difficult to do so. Nevertheless, for raw materials that have long lead times or highly variable demand, accurate forecasting is critical. This is the case with some raw materials that Bedoukian imports from international suppliers. These materials have both sporadic demand and a long lead time.

Since demand data on the utilization of raw materials are now available with the Chameleon ERP system, SmartForecasts can also be used to create appropriate lead-time forecasts for those items. Using Chameleon, the company tracks its usage of raw materials every day and compares on-hand inventory with minimum stocking level requirements on a weekly basis for replenishment purposes.

Bedoukian expects that the planning process will be even smoother now that SmartForecasts can be used to estimate optimal inventory stocking levels for its raw materials, many of which are used in multiple products. As Bedoukian moves forward with forecasting raw materials in the same way they forecast finished goods, they expect to have similar positive results and see more improvements in total supply chain efficiency and effectiveness.

The Results
“The intermittent demand forecasting capability has been very useful,” claims Eggleston. “We've gone from forecasting our slow-moving items by hand with poor results to automatically generating highly accurate stocking level estimates.”

In addition, SmartForecasts is helping the company see the patterns in its product demand so it can better track sales performance. This has been particularly useful for identifying changes in the buying patterns of customers, so that remedial action can be swiftly taken to satisfy these customers' needs and avoid losing their business.

Even though Bedoukian Research has been using SmartForecasts for almost twenty years, the company continues to increase its use of the software system and find new ways to benefit from its functionality. In fall 2004, Smart Software conducted a training and consultation session at Bedoukian at which representatives from production, sales, purchasing and management were present. The goal was for everyone to understand what SmartForecasts' brand of demand forecasting and planning could do for the company and how it works.

Before the session, Bedoukian had pretty much aimed for an overall 85 percent service level. And they had been achieving it! However, as a result of that session, Bedoukian began to prioritize its product inventory for demand planning purposes. For a few selected items that Bedoukian always wants to have available due to increased market competition, inventory and stocking requirements are estimated by SmartForecasts at a 90-95% service level. For the approximately 20% of their products that produce 80% of their annual sales revenue, inventory requirements are estimated at the 85% service level. For the remainder of their products, Bedoukian now uses lower service levels.

After just 6 months of using the new prioritization process, Bedoukian reduced its finished goods inventory by 5%, and the number of backordered items by 60%. The company now has more units in stock of the items they truly need to accomplish their business goals and fewer units of the items they do not need.

Bob Bedoukian is more than satisfied and claims, “The more we use SmartForecasts, the more important it becomes for our business and the more efficiently our business runs.”

Back to Success Stories