One extra tidbit from my interview with Murray Gaylord at The New York Times. In addition to his role as vice president for marketing, Gaylord also heads up the newspaper’s Consumer Insight Group, which studies Times Co. data, including print circulation and web analytics. He told me about a novel insight that has emerged from the group: One of their stats guys developed an algorithm to predict demand for tomorrow’s NYT by looking at today’s traffic to nytimes.com. Gaylord explained:

When there’s a big event that happens, and Spitzer is the example that we use, you get a big spike online, and then you know…you’re going to sell more copies at the newsstand the next day.

After the Times broke the Spitzer story online, their numbers guy said the Times should print 45,000 additional copies of the next day’s paper. “If you don’t print enough, you’re leaving money on the table,” Gaylord said, “and if you print too much, you’re losing money.” The Times hedged with 42,000 copies, which was about a thousand short of where the demand ultimately fell. Gaylord said the episode gave him confidence to use the algorithm for future big events like the election. I thought it was a good story because other penny-pinching newspapers might be able to copy the idea and make their expensive printing budget just a little more efficient.