The significance of Amazon having its highest month ever is minimal, as the world's largest e-commerce site has repeatedly broken its own record, as its traffic has grown over time. But the fact that Amazon saw more traffic in March—a traditional ho-hum traffic month—than December is noteworthy. To be fair, the Comscore figures show that March only barely beat December (105.3 million unique visitors for March 2013 versus 104.8 million for December 2012), but that marks the first time March has even come close. March 2012 saw 90.2 million versus December 2011's 97.4 million, and March 2011 had 61.5 million compared with December 2010 at 76.7 million, according to Comscore figures.
Bizarrely, one investor site report this week reported the opposite, that Amazon was experiencing a "severe slowdown in Web traffic" and attributed the figures to Comscore. Turns out the site had seen growth percentage figures (as happens to all large sites, growth percentages shrink as the numbers get huge) and confused them with raw traffic numbers.
The Amazon numbers are impressive because of a huge spike in traffic for March. Last month's 105.3 million compares with 89.9 million for February. Granted, February is a slightly shorter month, but the year before, that didn't make nearly that much of a difference. March 2012's 90.2 million compares with February 2012's 90.1 million. March 2011 saw 61.5 million, and February 2011 saw 60.8 million.
One of the age-old problems with Web analytics is that there are generally far too many variables to draw useful insights. Is March suddenly the new Über month? Or was this nothing more than a strong response to a lot of good promotions in March? Maybe good economic reports made consumers more comfortable with spending?
What if there's a yet unreported consumer trend to repeatedly switch browsers or devices? That could appear to transform lots of repeat visitors into far more unique visitors. Did one or two other sites have system problems in early March, temporarily pushing a lot of their visitors to Amazon? Maybe Amazon—or, for that matter, Comscore—changed it analytics slightly, but enough to no longer make December to March comparisons the so-called Web apples to Web apples?
Those caveats all neatly caveated, it's still enough of a pattern deviation to merit watching. Will we see this with other major e-tailers? More to the point, this wasn't a change in the steadily increasing patterns for December. The change was almost entirely March-based. Therefore, even if it becomes wide spread, it's not clear if any retail change would be merited, at least not until it becomes clear why the change is happening.
What if it turns out that early spring cleaning caused consumers to break a lot of household items that they now need to replace? Although, not so sure it merits much of a retail strategy change.