There’s no shortage at SAS’ analytics conference A2011 of discussions around statistical modeling and optimization techniques. Kevin Norwood from Procter & Gamble, for example, discussed his company’s use of clustering, optimization and data visualization to answer questions like, “How do we reformulate our detergents when faced with an ingredient shortage? How do we quantify the value of a new chemical? How do we optimize our portfolio of products to reach the desired intersection of performance and cost?”
Underlying this analytical brute force, however, is a deliberate shift in P&G’s human fabric: people are collaborating more. The traditional, "Let’s have a meeting to discuss the problem/send individuals off for independent analysis/meet again model," Norwood says, takes too long and enables individuals to advocate siloed solutions.
Instead, P&G colleagues collaborate in what’s called the Pixel Room: a room filled with eight high-resolution monitors that display data visually in a way that makes trends, relationships and anomalies come to life. This, according to Norwood, creates “a different way of working together,” allowing everyone to see both the forest and the trees. “You can see an enormous amount of data this way,” he said.
Norwood cited a specific example of discoveries that a team of modelers and marketers made while looking at sales data for a new product, “correlations they would not have seen if they hadn’t been together asking questions and interacting,” he said.