Managing Automation - December 2008 - (Page 23)

[ COVER STORY ] “With so many promotions going on, it’s difficult for CPG companies to forecast based on traditional historical information or techniques such as exponential smoothing,” says Danny Halim, vice president for supply management solutions at JDA Software, referring to statistical techniques to analyze time series data. “That’s not good enough anymore.” At food and beverage manufacturer Welch Foods Inc., for example, trade promotions account for about 15% of the company’s profits. But, says Welch’s CIO Ray Gosselin, the difficulty of predicting the success of a given trade promotion has also contributed to a 35% to 40% demand forecasting error rate and higherthan-desired finished goods inventories. Welch’s has gotten a handle on the problem by replacing spreadsheets and manual processes for tracking trade promotions with the trade promotions management module from Oracle’s Demantra demand management system. The tool, combined with a new demand planning organizational structure, helps Welch’s to model and predict the success and demand impact of promotions and has allowed the company to cut finished goods inventory by 10%, Gosselin says. But those aren’t the only factors contributing to demand forecasting complexity. As they attempt to boost growth and appeal to new markets, manufacturers such as P&G are supercharging their innovation processes, increasing the number of new products they are able to move into the market. As new customer type, style, color, etc. — to assemble a model of the product. Using that model, companies can forecast demand for the product based on historical demand for existing products with similar characteristics. Unfortunately, says Lora Cecere, an analyst at AMR Research, few manufacturers today are using attribute-based forecasting. That’s because many lack the skills or processes needed to develop salient product attributes, test them, validate them, and design databases that can be used to store and analyze attributes. THE PROBLEMS AT SMALLER COMPANIES This kind of complexity hits small and midsized manufacturers particularly hard. Bicycle parts manufacturer Quality Bicycle Products, for example, has seen the number of SKUs for which it must forecast demand grow gradually to 30,000 as its bike shop customers increasingly request unique products that they can private-label under their own brands. As product variety grew, QBP’s manual, WHERE THE APPS ARE spreadsheet-based demand planCurrent vs. planned supply chain application ning processes faltered. deployment (U.S.) based on a survey of 336 IT “Forecast quality was very low,” and line-of-business managers in process and says Jason Bates, supply chain discrete manufacturing companies. manager at QBP. “In fact, they Order Management were often so poor that the fore66% Current casts weren’t even used. Once a 21% vendor would look at how far off Planned they were, that would be the last Warehouse Management time they looked at it.” 58% Current The results, Bates says, 28% Planned were high inventory levManufacturing and Distribution Planning els and long lead times 53% Current from suppliers that didn’t 27% Planned trust QBP’s forecasts. Eventually, the $150 Transportation Management million company was forced to 52% Current bring in a statistical forecasting 33% Planned tool, Infor SCM Demand PlanDemand Management ning, and revamp its demand 52% Current planning organization, which had 31% Planned consisted of a single person hanInventory Optimization dling demand planning as well as 51% Current purchasing, order management, 35% catalog management, and cusPlanned tomer service management. QBP Service Parts Management and Planning created a dedicated inventor y/ 48% Current demand planning function re32% Planned por ting into the supply chain Supply Chain Network Design organization and has since been 47% Current gaining experience working with 36% Planned the Infor tool. The software allows the comSource: AMR Research pany to combine market intelli- inside look Using collaborative planning processes and tools, Linksys has cut inventory costs by 35%. — Linksys’ Payne products represent more of the mix, however, and as product lifecycles shrink, forecasting demand becomes more difficult. The problem, says Chris Russell, vice president for sales and service at supply chain design and optimization software vendor Optiant Inc., is that new products present no sales history for manufacturers to use to forecast demand by means of traditional methods. But there are new approaches that manufacturers can take to more accurately forecast demand even for new products. Supply chain software vendors, such as SAP, Oracle, i2, and Logility, offer demand planning products that work on a principle known as attribute-based forecasting. Rather than relying primarily on historical data, attribute-based forecasting looks at new-product features — selling price, Photo courtesy: Linksys 23 December 2008 ma

Table of Contents Feed for the Digital Edition of Managing Automation - December 2008

Managing Automation - December 2008 Contents Take 1 Business Objects Chief Says Union with SAP Meets Objectives After One Year Yes, Emerson, Too, Is in the MES Market Infor Chief Puts Off IPO, Restarts Buying Plans Kronos Now Tracks Shop Floor Machines IQMS Rolls Out User Interace, Other Upgrades Notes Five Ideas for Demand Planning Building on the SOA Blueprint Innovation Now A Team Effort Lean %2B Technology = LEAN^2 Finding Flaws Before They Spread Product Scan Advertiser Index Next

Managing Automation - December 2008

For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.