Companies that have mastered demand forecasting and effectively use that information to drive their business operations have the best supply chains. At least that is the premise that Boston-based AMR Research has used to develop its list of top 25 supply chains.

According to Kevin O'Marah, vice president of research at AMR and lead analyst for its supply-chain top 25, cutting-edge companies in the last five years have reconfigured their supply chains and their approach to supply-chain management around using demand signals to drive their businesses.

"We all have known that it is better to pull than to push demand," says O'Marah. "The 25 companies on our list actually do it."

Companies on the AMR list such as P&G and Dell have completely changed the way they use point-of-sale information, make forecasts and fulfill orders. According to O'Marah, these companies capture and manage demand dynamically, and then use the information to position inventory, plan manufacturing capacity and execute orders.

"Good supply-chain management is more than low cost and efficiency," he says. "It requires a superior ability to shape and respond to shifts in demand with innovative products and services."

This new model is what AMR calls demand driven supply networks (DDSN), where companies drive growth through constant interplay of supply, demand and product domains.

While many companies brag about their supply-chain performance, O'Marah says that there is a measurement void that makes valid comparisons difficult. To fill this gap, AMR has developed its list of top supply chains based on a combination of subjective opinions from an anonymous board of industry experts and from publicly available financial results. The opinion portion counted for 40 percent of the index and the financial results for 60 percent.

By the Numbers
The financial results used include such metrics as return on assets, inventory turns and trailing 12-month growth. These metrics were combined with the opinions of a panel AMR Research put together to evaluate each company's supply-chain performance. The result is the composite ranking score for 25 companies on the accompanying table (PDF).

O'Marah says that the financial numbers alone cannot be used to compare supply chains. Return on assets (ROA) does not mean the same thing to two different companies in terms of their operations, but he believes ROA and other financials provide important supply-chain insights.

"If a company has $100m in assets, I want to know what the return is," says O'Marah. "We need a baseline for how effectively dollar assets are deployed."

All three of the financial measures that AMR uses are publicly available and help gauge how well a supply chain is performing, at least compared to like companies. AMR collected these financials for all of the companies considered and broke the companies out into industry groups, so the panelist could look at competing companies.

Of course, all the financials are just a snapshot in time. ROA, for example, is based on the end-of-a-year asset valuation.

"We are not really concerned about which companies did well in the past," says O'Marah, adding that these metrics on inventory, return on assets and growth provide a prospective view of which companies will be doing well next year and the year after.

O'Marah says that even the opinion portion of the index is partially based on objective metrics. AMR put together an anonymous panel of 12 industry and functional experts. The 12 panelists selected the 25 names from a list of 250 global manufacturers and retailers based on how well they have adopted DDSN principles to run their supply chains and grow their businesses.

The panelists were given strict voting criteria based on the formal definition of the DDSN "orchestrator." AMR defines the DDSN orchestrator as the company that manages demand proactively throughout an extended supply chain and has the power to reconfigure the supply chain as necessary. Based on their own knowledge of these companies, the panelists selected which companies best met the definition from a list of the 250 manufacturers or retailers in the global 500. The selection process consisted of several steps. The panelists first picked between 30 and 50 companies from the list of 250 candidates. O'Marah combined all the panelists' selections into a finalist list of about 40 companies. The participants then picked and ranked the final 25. O'Marah merged the final lists into the AMR opinion component for the Top 25. This list counted 40 percent of the final number.

"This is an opinion, but it is based on our knowledge of the supply-chain leaders and how they operate," says O'Marah.

Many of the companies on the AMR list are either retailers or original equipment manufacturers (OEM). As such, they are often in the best position in the supply chain to be the orchestrator because they are closest to end-user demand. Orchestrators can also be suppliers. For example, Intel is a supplier up the supply chain from its OEM customers, but in many supply chains, it is the orchestrator that manages demand for its customers.

For example, O'Marah explains that Intel's Centrino chip created a market based on a technology only it could bring to market. Intel embraced the demand variability for this chip as an opportunity and used the demand to run the supply chain for its OEM customers. Similarly, Korean steel producer POSCO is an orchestrator even though its material supplies are way upstream for its manufacturer customers.

"POSCO makes the list because of the sophisticated way in which they leverage their investment in their production assets, and because they are able to manipulate demand for the finished materials," he says. "They don't just build a steel mill and hope to sell the steel. They build a plant knowing a great deal about the downstream demand."

Key Metrics
According to O'Marah, four key metrics were used to determine how well companies use DDSN to run their supply chains and were factored into the opinion portion of the index:
• Demand forecasting accuracy
• Perfect order fulfillment (complete, on time and accurate)
• Supply-chain cost
• Cash-to-cash cycle time
Demand forecasting accuracy is the key metric that supports the other three, according to O'Marah. For example, companies that excel at demand forecasting average 15 percent less inventory, 17 percent better order fulfillment, 35 percent shorter cycle time and 90 percent fewer stockouts than companies with poor demand forecasting accuracy.

According to O'Marah, demand forecasting accuracy also correlates with perfect order fulfillment. A one-percent improvement in demand forecasting accuracy yields a two-percent improvement in perfect order fulfillment.

	"Good supply chain management requires superior ability to shape and respond to shifts in demand with innovative products and services." - Kevin O'Marah of AMR Research

Subsequently, perfect order fulfillment performance results in a 50-percent improvement in earnings per share. A five-percent point improvement produces a 2.5 percent improvement in return on assets, and a three-percentage point improvement in perfect order fulfillment correlates with a one-percent boost in profit margin.

The measurement of these underlying metrics of supply-chain operational performance is the measurement of the business's efficiency and is definitive to its business success. By mapping companies on a graph based on these operational metrics along with a company's product innovation capability (time-to-value), industry leaders all appear at the high, right end of the graph.

"The combination of these attributes gives a good picture of which companies will continue to have both ongoing profitability and good prospects for future growth," says O'Marah.

IBM is one company harnessing the benefits of supply-chain best practices. In 2002, it began to tie all the pieces of a supply chain together. This required managing over 30,000 suppliers, offering 78,000 products and operating 13 manufacturing plants in 9 different countries. IBM transformed its supply chain from a cost of doing business into a competitive advantage.

The results of the transformation have made a significant impact on IBM's business. Over the last two years, IBM has reduced cost and expense by more than $12bn. Inventory levels are at a 30-year low while the amount of sales is increasing. IBM has increased intra-division part reuse from two percent to over 50 percent and annual part number reductions range from 20 percent to 38 percent.

"In the on-demand era, the real opportunity is about how you design a business and all its supplier relationships in ways that really haven't been possible until now," said Bob Moffat, senior vice president, Integrated Supply Chain. "Advantage will flow to companies that embrace this new model and demonstrate the ability to simultaneously excel at the fundamentals of supply chain and leverage next-generation management practices."

Factoring in Technology
Technology and internet capability were ultimately factors in which companies made the list, but not just because the company invested in the latest supply-chain software.

According to O'Marah, high investment in enabling internet technology helped the companies on the AMR list to proactively manage their supply-chain operations.

"Their technology investment is done in conjunction with best practices, especially those focused on sales and operational planning, lean manufacturing and vendor managed inventory," he says. "Only technology spending targeted to specific operational problems and daily use of best practices leads to success."

O'Marah points out that it is impossible to run a global demand driven supply network without technology.

"People love to say that supply-chain performance is about business process," says O'Marah. "That is not entirely true. It requires really good technology as well."

In fact, all of the companies on the list use off-the-shelf supply-chain solutions from vendors such as SAP, i2, Manugistics, and others. They also all invest in custom-built supply chain systems.

According to O'Marah, the top 25 supply-chain companies on the AMR list have three things in common, all of which are partially depended on technology:

• They proactively manage demand, which means that they gather multiple demand signals. They do not work off of an arbitrary single number forecast. They are comfortable with demand variability and treat it as an advantage. O'Marah points to Dell as the perfect example. Dell raises prices on high demand items in short supply. If they are don't have it at all, Dell hides the item from view on their web site. If Dell has too much of an item, they will discount it. They constantly adjust pricing against stock, so demand variability becomes an opportunity.

• They embed product innovation in their supply chain. These leading companies do not treat product innovation as an externality. R&D and engineering doesn't just throw a product over the wall and then have manufacturing and sales figure out how to make money on it. They have a comprehensive feedback loop between what is selling, what is working in the market place, how it is performing in the field and the next generation of product needs to look like. They design for the supply chain.

• They use probabilistic demand forecasting solutions. Instead of relying on supply-chain optimization solutions that are deterministic, i.e., use linear programming to find one optimal answer, they use optimization algorithms that are probabilistic and that quantify the range of possibilities around a demand forecast or supply capacity. They model demand for that variability.

"The basis of competition for winning companies in today's economy is supply-chain superiority," says O'Marah. "The companies on our list understand that value chain performance translates into productivity and market share leadership."