Widget Metrics and Linguistic Precision


Comscore today announced the launch of “comScore Widget Metrix,” a new service to track the usage of widgets across the Web. The day before this announcement, VentureBeat posted the news that according to comScore’s numbers the slideshow company Slide is the world’s top widget provider, with the title “Slide pounds chest: Widget used by 14 percent of internet population.”

The problem is not that comScore is applying metrics to widgets, nor that Slide is extremely popular. The problem is that the comScore report very clearly uses “unique visitors” as its unit, but VentureBeat shifted that metric to “users;” however you may choose to define “using” a widget, loading a Web page that happens to have some embedded flash certainly isn’t it.

This kind of analytical sloppiness is what makes me suspect that there’s a widget bubble building up in the first place. If it is accurate to say that 14% of the Internet “used” Slide, it is also accurate to say that about 90% of the Internet “used” DoubleClick in the same period: in both cases we’re talking about something being displayed for an Internet user who may or may not have even noticed it was there.

Because widgets are new and cool, we want to believe that widgets are genuinely a different animal, and that they won’t — couldn’t — end up in the roadside ditch where we dumped banner ads. It’s easy, right now, to make the slip of reading comScore’s report and thinking about all those people using Slide’s widget, without ever realizing that you just transformed passive viewers into active users.

But before the Widget Mafia starts sending me angry notes again, let me make one thing clear: I absolutely believe that some widgets are genuinely a different animal, and that those widgets are going to change some of how we think about and use the Web.

I also believe a couple of other things, though. I believe that while “unique viewers” is a better metric than none at all, the widgets that matter are going to need something very, very different: be it actual widget “usage,” “engagement,” or what have you, widgets get interesting when people want to interact with them. And yes, I believe that the overwhelming majority of the widgets that we see in coming months and years will not be any more interesting than punch-the-monkey banner ads, and that if we want to avoid a widget bubble we need to maintain a healthy degree of skepticism.