Q&A With James G. Robinson, Director, Web Analytics, The New York Times

At The New York Times, Web analytics is at the core of its print strategy, not just online.

By Chandra Johnson-Greene
Wednesday, January 14, 2009

James G. Robinson joined The New York Times in October 2007 as its Web analytics director, when the company was implementing a strategy to integrate the newsroom, sales and Web departments. The Web analytics group is part of the larger Customer Insights group, which also includes the customer research and Web user experience groups, and contains employees with a diverse and valuable group of skill sets, Robinson says, including those who have worked in circulation and database marketing.

Describe your relationship with the rest of the staff. How do you package custom reports for the sales, editorial and corporate groups, for example?

We have a variety of tools we use to get the word out. We produce regular reports for the newsroom, sales and marketing teams that include weekly and monthly traffic results. We also provide special request reports. But we do keep away form sort of throwing numbers out there because it’s not very useful. If we just give them a large spreadsheet and they don’t read it, then it’s not actionable.

But beyond the dashboard, we also provide a single page bulletin about once a week that very clearly conveys one insight or idea—something interesting that we think the rest of the company should be aware of. It’s sort of our version of breaking news.

Describe how NYT’s strategy is influenced by Web analytics and what business decisions are based on them.

There’s a strategy behind everything that we do. We don’t like being in the trivia business. All of our information is used to make better decisions. For example, the biggest story we broke this year was the Elliot Spitzer scandal. We broke it first online at 2 pm on a Monday. By 4 pm, it was clear our site traffic was going to be through the roof.

Four o’clock is a really interesting time here because that’s when the senior editors meet to decide what the top story will be for the next day as well as when the circulation team decides how many papers to put out on the street. We came up with the idea to go back to the past few years and look to see whether unexpected news events that broke online caused a spike in newsstand sales the next day. We found out that there was a correlation, so we were able to make a better guess at how many papers to put on the street. It was a direct bottom line impact.

What are the most valuable metrics to the NYT? And has that changed at all over the last year?

The metrics we look at depend on the specific business need at hand. Time spent, for example, is a key metric when we serve ads on the site that are based on the number of minutes spent on that section. But for us, there’s no magic metric that will explain what the driving force is behind everything that we do.