Voice of Experience: Purchase Manager Saves With AuctionsBy Baselinemag | Posted 2006-01-14 Email Print
Re-Thinking HR: What Every CIO Needs to Know About Tomorrow's Workforce REGISTER >
Jim Polak, director of purchasing at PPG Industries, says online auctions will save his company at least $10 million per year.
MANAGER'S PROFILE: Oversees about $1.3 billion in annual spending on goods and services. PPG, with sales of $9.5 billion in 2004, makes paints, adhesives, glass and industrial chemicals.
HIS PROJECT: For the past two years, PPG's 150 buyers have used Ariba's spend analysis software and its hosted sourcing tools to solicit bids from suppliers. The company originally used e-sourcing tools from FreeMarkets, which Ariba bought in 2004.
FOLLOW THE MONEY: Polak knew PPG could save money with online auctions and other sourcing tools, but "the first thing we needed to do is find out where the dollars were going," he says.
SAVE WAVE: By analyzing PPG's spending, Polak and his team identified commodities, including temporary labor and manufacturing equipment parts, that were ripe for auctioning to single suppliers. Over a three-year period, the company expects to save more than $10 million per year by putting those categories up for competitive bids. While that's less than 1% of PPG's total spending, Polak notes that prices have increased in many categories at the same time.
LOWBALL FIGHT: Suppliers can be squeezed only so far, Polak says. For example, in 2002, PPG held an auction for electrical maintenance, repair and operations services and saved 20% compared with what it was previously paying. When it held the same auction in 2004, the winning bid was just 7% lower than the previous contract.
MESSY'S OK: Polak says product or vendor data doesn't need to be normalized before an online auction. For example, he says, a 3-inch valve from vendor A may have a different purchase-order code associated with it than the same type of valve from vendor B. But that's OKthe buyer overseeing the auction just needs to know that the products are comparable. As Polak puts it: "You don't have to wait until your data is perfect."