Like a Well-Oiled Machine

Lubrizol Doubles Accuracy of Demand Forecasts with SAS®

The Lubrizol Corporation likes to say, "Our vision for the future is based on the most valuable lessons of our past." That philosophy continues to make the 75-year-old Cleveland, Ohio-based fluid technology company successful. With $2 billion in annual revenue, Lubrizol is a recognized leader in specialty additives for lubricating oils. Their lubricant and fuel technologies increase operating efficiency and reduce the environmental impact of all types of transportation equipment.

While Lubrizol makes products that keep the world running smoothly, they are striving to have their own operations run smoothly and efficiently thanks in part to a forecasting system built with SAS software.

"Demand planning and forecasting impact just about everybody at Lubrizol, including our finance, sales, marketing and manufacturing groups," says Robert Wojewodka, Lubrizol's Operations Technology Manager and Statistician. "It impacts our financial group because they need to know where we are, where we are going, and what our expectations are for the coming year. At the other end of the spectrum is manufacturing, where we must figure out what we need to make and when we need to make it."

Choosing SAS an 'Easy Decision'
Lubrizol serves more than 2,000 customers in more than 100 countries on five continents – which adds up to over 30,000 unique forecasts each month. Account managers need forecasts for each of these customers by product and location, which, before SAS, was a daunting task. Lubrizol trudged through manual activities that often relied on subjective "gut feelings" and daily planning meetings at individual facilities. Given the complexity of forecasting and the lack of tools to support the process, the quality of the forecasts varied widely and forecasts often were not updated.

From analysis of their historical data, Lubrizol realized that its forecasts could be improved. Incorrect forecasts, forecasts allocated to the incorrect customer locations, and forecast biases were present. Incorrect forecasts rippled back through the supply chain, creating inefficiencies in logistics, production and materials sourcing.

So the company looked for a fast, accurate way to do the job and assembled a competent and energetic team to get it done. Given the dynamic nature of the business, Lubrizol wanted a solution that was very powerful as well as easy to use and customizable. "Lubrizol has used SAS in a variety of areas for over 15 years, including research and development, business analysis, and operations analysis," Wojewodka says. "And since we knew its capabilities, selecting SAS as the analysis engine for this type of project was an easy decision to make."

Automated System Conquers Forecasting Volume
Using SAS, Lubrizol has been able to create a more automated forecasting process. SAS brings together historical data spread across multiple systems, including SAP, Excel and Access, and automatically generates statistical forecasts for the next 26 months, taking into account the latest trends and patterns in the demand. SAS then applies additional business rules that Lubrizol has determined are important, providing even more flexibility and intelligence to the system

As a result of this automation, "we are making life easier for 120 account managers," Wojewodka says. "We have taken a lot of the workload off them and shifted it so they can concentrate on more important things. The account managers thank us for that, and senior management responds accordingly."

"Forecasts that used to take a very long time to generate or were not generated at all are now completed in minutes," he adds. "Because of the increased automation, account managers can also focus on supplying more important information to the forecasting process. The account managers are being asked to add their knowledge of customer-specific trends and activities to adjust the statistically generated forecasts. They are being asked to concentrate on obtaining and inputting information that a model does not have. Is a customer going to be moving purchases from product A to product B? Is the customer planning to shift volumes up or down or moving shipments to a new production location? These are examples where our account managers provide great value in adjusting a model based forecast. The system, in return, provides a variety of reports, graphical summaries, and key performance indicators to support both the sales and operational activities."

"Account managers are now encouraged to review big ticket items first based upon selected criteria and let the statistical models work for everything else," says Cherie Sutherland, Lubrizol's business information manager. "Now they proactively review key variance reports as opposed to spending time doing lots of updates."

"Our account managers are starting to use the summaries and graphics to talk with our customers about what the future holds, which is something we couldn't do very easily before SAS," Wojewodka says. "Our intent is to start using this information more and more within our manufacturing plants to sequence production and raw-material handling. Ultimately, purchasing may further use the information to make sure we have the materials we need to meet demands. And our planning and support group uses economic information from our forecasts to determine at which location we can manufacture products while keeping costs at a minimum."

'You've Got New Sales'
According to Wojewodka, a unique benefit of how Lubrizol is using SAS is how Lubrizol has set it up to discover new sales situations that have never occurred before. When it spots something new going on, SAS automatically builds a report that presents the situation and lists the customer, product and location. SAS then e-mails the report to the appropriate account manager under the subject header: "You've got new sales."

"This allows the account managers to start a dialog with the customer and receive some initial forecasting," Wojewodka says. "We receive a lot of compliments from our account managers on that. We have even received feedback from several customers, because very often we are uncovering product or purchasing trends that even they weren't aware of."

Lubrizol is also using SAS to process many business rules that they have developed when generating demand forecasts. For example, their system detects data about products that have been scheduled for obsolescence and adjusts forecasts accordingly. In the past, sales managers often sold products without realizing they were being phased out, which prompted inefficiencies in the sales and manufacturing processes. With SAS helping to detect and plan for obsolete products, Lubrizol can now manage business relationships around products more effectively than before. "The ability to easily combine business rules in conjunction with rigorous statistical analysis while staying in the same platform is a strong feature of SAS," says Wojewodka.

More-Than-Doubled Accuracy Pays Off
Since implementing their solution, Lubrizol's forecast accuracy has more than doubled. "This is very good news considering that each one percent reduction in forecast variance equals $200,000 in reduced expenses as identified by our planning and support group," Wojewodka says. "SAS is giving us a better handle on our future requirements, which allows us to be more cost-effective. With SAS' capabilities, we can now have a better dialog with our customers and can show them the demand we anticipate them having for our products. This improves our relationships with our customers because we become true business partners, and we manage costs and improve customer service in a way that's mutually beneficial."

"However, everything is relative to the starting point," Wojewodka adds. "Although we have doubled the forecast accuracy from our baseline, we still need to make further improvements. This will come from further improvements to the modeling process along with more effective use and input from our account managers and our customers not to mention our newly formed demand planning team. It takes everyone working effectively together to really make a difference."

"Ultimately, SAS' strength lies in its speed and ease of use," Wojewodka says. "What is really beneficial is the power, speed and flexibility," he says. "Within SAS you can extract and process data to organize and preprocess it, and you can analyze and report it in HTML or standard reports, XML, text and graphics. It has incredible power as an analysis and forecasting engine. Within one platform, you can do almost everything. I think that's very powerful."

Concludes Ellen Rosen, Lubrizol's senior information technologist: "With SAS, we are able to do some very complicated data manipulation with relatively little code. It is also fairly simple to produce professional-looking, real-time graphics and use these graphics on a Web interface."

Copyright © SAS Institute Inc. All Rights Reserved.