Unilever Gets a Better Sense of Demand
Even before the volatile global economy and unsteady consumer spending, Unilever was on task to decrease costs and maintain profits. In 2006, as part of a company-wide initiative to improve forecasts, Unilever North America tapped Terra Technology to pilot its Demand Sensing solution.
“We’re always trying to improve our forecasts, which drives downstream benefits to the business from lowering costs and improving service to reducing our carbon footprint,” says Doug Sloan, director of Supply Chain Operations Support, Unilever North America. “As part of that broader initiative we discovered Terra Technology, which had a unique capability to contribute significantly to improving tactical forecasts and thus improving a lot of those other measures.”
Using a year’s worth of Unilever’s data, the tool applied pattern recognition algorithms to downstream data to create an accurate prediction of demand. Results were encouraging, but an internal business integration project stalled implementation. Once back on track, it deployed Demand Sensing in 2009, starting with its U.S. retail business.
The software bolts on to Unilever’s existing demand planning systems, a factor that allowed for a smooth implementation with no disruption to systems or processes.
“One of the things we liked most about the tool is that it’s touchless,” says Sloan. “It’s quite different than many other applications that require significant process change, change management work or systems integration. The forecasting person still does the work he or she did before, but Terra takes the raw data and further optimizes it for the one-month horizon, which frees up time for them to focus on longer term forecasts.”
Some work was necessary to integrate Demand Sensing with its legacy inventory optimization system, which is now being replaced by another Terra tool, Multi-Enterprise Inventory Optimization.
A little more than a year after going live in North America, the company has decreased forecast error in manufacturing and inventory. It realized a 40 percent improvement in forecasting accuracy for the seven-day horizon and a 16 percent improvement in the 28-day horizon across all of its brands.
“That translates to three days of safety stock improvement using the legacy inventory optimization tool,” says Sloan. “We also improved freight costs and we know it has improved our carbon footprint primarily because of reduced freight and maintaining lower inventory levels in our warehouses.”
Based on the results in North America, there is an active project underway to roll out demand sensing and inventory optimization tools simultaneously in key markets globally. In fact, Unilever anticipates significant safety stock reductions following the worldwide implementation.
Unilever is also exploring the next generation of Demand Sensing, Multi-Enterprise Demand Sensing, which reconciles multiple daily demand signals from more echelons in the supply chain, including point-of-sale data, channel inventory and retailer forecasts.
FAST FACTS
Optimizing Inventory
Unilever North America is now in the process of deploying Terra Technology’s Multi-Enterprise Inventory Optimization to optimize inventory across all echelons of the supply chain.
Company at a Glance
Unilever, whose global revenue totaled about $57 billion last year, has 270 brands in 13 categories that are marketed in 200 countries. Eleven of those brands generate more than $1 billion in sales annually.
Words of Wisdom
"Continuous improvement of forecast accuracy is one of the key enablers of Unilever’s supply chain strategy to lower costs, improve cash delivery and deliver excellent customer service." — Doug Sloan, Director of Supply Chain Operations Support, Unilever N.A.
Going Overseas
A global team was formed to pilot and deploy the demand sensing and inventory optimization solutions for Unilever in key markets around the world.
“We’re always trying to improve our forecasts, which drives downstream benefits to the business from lowering costs and improving service to reducing our carbon footprint,” says Doug Sloan, director of Supply Chain Operations Support, Unilever North America. “As part of that broader initiative we discovered Terra Technology, which had a unique capability to contribute significantly to improving tactical forecasts and thus improving a lot of those other measures.”
Using a year’s worth of Unilever’s data, the tool applied pattern recognition algorithms to downstream data to create an accurate prediction of demand. Results were encouraging, but an internal business integration project stalled implementation. Once back on track, it deployed Demand Sensing in 2009, starting with its U.S. retail business.
The software bolts on to Unilever’s existing demand planning systems, a factor that allowed for a smooth implementation with no disruption to systems or processes.
“One of the things we liked most about the tool is that it’s touchless,” says Sloan. “It’s quite different than many other applications that require significant process change, change management work or systems integration. The forecasting person still does the work he or she did before, but Terra takes the raw data and further optimizes it for the one-month horizon, which frees up time for them to focus on longer term forecasts.”
Some work was necessary to integrate Demand Sensing with its legacy inventory optimization system, which is now being replaced by another Terra tool, Multi-Enterprise Inventory Optimization.
A little more than a year after going live in North America, the company has decreased forecast error in manufacturing and inventory. It realized a 40 percent improvement in forecasting accuracy for the seven-day horizon and a 16 percent improvement in the 28-day horizon across all of its brands.
“That translates to three days of safety stock improvement using the legacy inventory optimization tool,” says Sloan. “We also improved freight costs and we know it has improved our carbon footprint primarily because of reduced freight and maintaining lower inventory levels in our warehouses.”
Based on the results in North America, there is an active project underway to roll out demand sensing and inventory optimization tools simultaneously in key markets globally. In fact, Unilever anticipates significant safety stock reductions following the worldwide implementation.
Unilever is also exploring the next generation of Demand Sensing, Multi-Enterprise Demand Sensing, which reconciles multiple daily demand signals from more echelons in the supply chain, including point-of-sale data, channel inventory and retailer forecasts.
FAST FACTS
Optimizing Inventory
Unilever North America is now in the process of deploying Terra Technology’s Multi-Enterprise Inventory Optimization to optimize inventory across all echelons of the supply chain.
Company at a Glance
Unilever, whose global revenue totaled about $57 billion last year, has 270 brands in 13 categories that are marketed in 200 countries. Eleven of those brands generate more than $1 billion in sales annually.
Words of Wisdom
"Continuous improvement of forecast accuracy is one of the key enablers of Unilever’s supply chain strategy to lower costs, improve cash delivery and deliver excellent customer service." — Doug Sloan, Director of Supply Chain Operations Support, Unilever N.A.
Going Overseas
A global team was formed to pilot and deploy the demand sensing and inventory optimization solutions for Unilever in key markets around the world.