The Problem With Late Research, Or No Research At All

comScore, which measures online consumer behavior, is among the best companies at analyzing traffic to websites.  comScore also decodes what people do when they get to their online destinations. The value of the data is fleeting, though. The internet is forever changing as the habits of people and businesses using the web move in new directions every day. Some websites lose popularity while others gain it.

The new comScore data about e-commerce is nearly useless, not because it was poorly collected, but because it is old. The information shows e-commerce activity in November and compares the figures to October. The trouble is that Christmas has now passed and online shopping habits are ancient history. There were large surges in visits to the e-commerce sites for Best Buy (NYSE:BBY), Sears (NASDAQ:SHLD), Wal-Mart (NYSE:WMT), and Toys “R” Us in November. That does not matter at all on December 28 when the information was released. It is already clear that Amazon (NASDAQ:AMZN) had the most holiday traffic in the online shopping world. A number of research firms determined that last week.

The comScore data about big e-commerce websites in November also does not show what people bought when they visited the sites, whether they came back again, or whether they liked or disliked the websites when they visited them. Merely saying that some people visited a given site during a specified period does not give much valuable information. This data is useless except as an idea for a headline. It might read “Best Buy Website Has 67% Increase In Visits During November.” All of those site visitors may have looked at the Best Buy home page, may not have liked what they saw, and then moved to Amazon.com.

The comScore data is the kind of information the press gets from many research organizations today. It is “topline” data that shows a summary of trends that are, usually, terribly complex and in need of interpretation by experts. It almost never turns out that way. Reporters and editors take a few sentences and declare some few pieces of data show an important trend. The information used to create the numbers used to create the headline may show something completely different. But, the media usually does not ask for that more detailed data or neglects to check with an expert about what the numbers mean.

The biggest headline of the day yesterday was that holiday spending rose 3.6%, according to data from a division of MasterCard. Somewhere among the footnotes was that there was one more shopping day this holiday season compared to last year. Information from the MasterCard poll is inadequate. It measures activity from its credit cards but estimates purchases made by cash or check. Those estimates may be off considerably. The National Retail Federation affirmed its estimate that retail sales would drop 1% for the season. It was the association’s way of saying the MasterCard data was wrong. A percentage point means a lot when the amount at stake is in the tens of billions of dollars.

The temptation to take a headline based on the most modest information and run it is probably as old as the press is. There is certainly no sign that this behavior will change anytime soon. It sells newspapers and brings eyeballs to news websites. This habit does not do the news consumer much good. He is left with an impression about some part of the world of business, or sports or entertainment that is barely true, and the press knows it. How else could website TMZ run a picture of JFK on a boat with naked women from fifty years ago, only to admit the next day that it was fake?

Douglas A. McIntyre