One of the basic tenets of business is balancing supply and demand. For the past several years, manufacturers have focused on better managing the supply side of the equation. The emphasis on upstream activities was natural because of a strong economy and the need to create more agile supply chains. Now that the economy has changed, organizations that have worked to get their supply chains in order are looking to better understand and manage their demand signals and forecasts so they can make the right products, in the right places, in the right amounts to meet demand and not have excess safety inventory tying up assets.
For companies like Bandag Inc. (Muscatine, IA), makers of equipment and materials for re-treading tires for the trucking industry, demand management means "internally understanding the marketplace and [knowing] when product is going to be requested and in what quantity throughout the markets we serve," says Greg Harris, Bandag's manager of production planning and logistics.
"There's been a shift to a demand-driven enterprise. Companies have been focused on getting rid of inefficiencies in their supply chain but now that the buyer has been empowered by the Internet, you're seeing an uptake in demand-chain management," says Kent Allen, research director of demand chain management with Aberdeen Group Inc. (Boston, MA).
For most manufacturers, getting more accurate data to create better forecasts is the key to not building the wrong product or stocking a product that isn't going to sell. Today's organizations also need to be able to adjust forecasts more frequently. This focus on demand management will contribute to a collaborative planning and scheduling software market that is expected to grow by 25% annually from $2.9 billion in 2002 to $5.6 billion by 2005, according to ARC Advisory Group Inc. (Dedham, MA).
Software providers including ERP vendors Oracle Corp. (Redwood Shores, CA), PeopleSoft Inc. (Pleasanton, CA), and SAP America (Walldorf, Germany) offer demand forecasting tools. Supply and demand chain vendors like Demand Management Inc. (St. Louis, MO), i2 Technologies (Dallas, TX), Manugistics Group Inc. (Rockville, MD), Haht Commerce Inc. (Raleigh, NC), Prescient Systems Inc. (West Chester, PA), and Syncra Systems Inc. (Waltham, MA) also offer such tools. These tools provide sophisticated algorithms to help companies develop forecasts based on both historical and real-time data from customers and distribution channels. They also provide Web interfaces for manufacturers to gain more visibility into demand signals as well as collaborate with their customers and channel partners. It's no surprise that these software suppliers also provide supply chain applications since demand has to be balanced against inventory and supply chain production capabilities.
Accurate forecasts help reduce inventory in the chain, increase profitability, and manage the lifecycle of products, but are manufacturers ready to collaborate with customers and partners on those forecasts? For most companies, better forecasting starts with improving internal processes so forecasts to suppliers are more accurate before exposing those processes to the rest of the world. Better demand management often means looking at a lot of variables. It often means combining historical data with current market conditions, ongoing and planned promotions, available inventory, and supply chain information. Much of this information can only be gathered if there are open lines of communication with customers, distribution channels, and supply chain partners.
"The new collaborative tools help companies share forecast information, but, more than anything, these tools encourage better communication between partners on demand forecasts which creates a better understanding of all the factors involved," says Aberdeen Group's Allen.
There are also potential pitfalls to collaborating with customers on demand forecasts. Other models might be more effective as these tools emerge and mature. "The customer you collaborate with and allow to change your forecast may be the same person who's been ordering the wrong products for years," says Mike Campbell, president and CEO of Demand Management. Campbell says a better process for most companies is to provide the forecast to sales and marketing, or even customer service, and have those people meet with customers to get consensus on the forecast. Not only will there be a greater understanding of what influences the forecast, but the internal person will also have more information about the product.
"There's a problem with too much emphasis on the demand side," Campbell notes. "If you look at the entire supply chain, there are a lot more elements involved than just the forecast. More value, given that the forecast is wrong, comes from taking the forecast and melding that with the information in the rest of the chain to find out where you really stand."
Bandag uses Manugistics' applications to forecast demand, oversee inventory and production planning, and manage outbound logistics. To perform better demand forecasting, the company first focused on "understanding how to forecast our products for suppliers," says Harris. Bandag benefits from allowing its applications to take the data they have and, using the algorithms provided, let the applications come up with the best forecast and plans. This is possible because the company's product line is relatively stable and the demand planning process "doesn't require a lot of judgment events. The real trick is having the right amount of product available," says Harris.
Bandag's implementation began two years ago with the goal to better understand demand to reduce back orders and better manage inventory. The only way to do that, says Harris, "is to have accurate forecasts and be comfortable that the forecast is going to perform going forward." Using Manugistics' forecasting module to monitor inventory, Bandag reduced back orders by 50% the first year and another 44% the following year. With that history in place, the company was able to reduce inventory by 10% so far this year as well as even out its production activity. Prior to the Manugistics implementation, the company built up and stored inventory in non-peak times for use during peak times. "It requires you to take a hard look internally at your systems and processes," says Harris.
The company does provide forecast information to its salespeople, and encourages them to share forecast data with customers. But Bandag isn't yet encouraging salespeople to provide feedback on the forecasts because the company found it got more variations from other inputs to the forecast than it had just letting the system statistically forecast demand based on historical data. Based on that experience, it will be some time before the company will consider Web collaboration with customers.
Other companies are taking a more collaborative approach both internally and externally, but are not yet at a point at which collaboration and adjustments to forecasts are automated.
SSL International plc (Cheshire, UK) manufactures medical and consumer products such as surgical gloves, wound dressings, and foot-care products. Like Bandag, SSL uses its PeopleSoft implementation to run historical data through statistical algorithms to come up with a suggested forecast. With that forecast in hand, internal sales and marketing, logistics, production, and financial representatives meet to discuss the forecast. When they all agree, that forecast is plugged into the company's sales and operations planning (S & OP) process and put into action.
With large customers like Wal-Mart, SSL takes a more collaborative approach. Each company creates a forecast and then they exchange their numbers. "Then we meet to talk about the differences in the forecasts on a quarterly basis and work to lessen any gaps," says Jim Hourigan, vice president of IT and logistics with SSL.
SSL talks with several customers each quarter to get more specific information, and reviews and adjusts its internal forecasts monthly. The company also exchanges inventory and point-of-sale (POS) data with some distributors via EDI to get more current data to plug into the system. Before using those numbers to adjust forecasts, Bandag meets with its distributors to get the story behind the numbers. "You have to have the baseline numbers to talk to but it's really important to put the story around the numbers so you can understand the issues," says Hourigan. "The PeopleSoft forecasting and planning applications are the enabling software that lets us spend more time talking with our customers. That's where you get the big payoff as it strengthens the relationship," he continues. "The more you do that and the stronger the relationship is, it almost becomes a barrier to entry for the competition."
Demand management and forecasting applications need to be integrated with other systems, such as ERP systems, to access some of the historical data that's needed to create plans; integration with supply chain systems provides a view of production capabilities and other issues that may influence the performance of the forecast. As companies strive to get more demand information from their downstream partners, these applications will need to be integrated with customer relationship management (CRM) applications. "There's a lot of customer information and behavior [that] is captured by CRM solutions. The Web provides a way for companies to implicitly and explicitly get information about customers," says Aberdeen's Allen.
Even with these advanced forecasting tools, manufacturers and others in the supply chain still rely on good planners and "the human aspect of being really good at the art and strategy of planning," says Tim Douglas, senior marketing manager of SCM with Oracle. So expect demand forecasting to continue to be more art than science, with the processes supported by improving tools and integration. MA