The Web Is Supposed to Be About Metrics, Only Whoops, You Can't Measure Anything

By Steve Bryant  |  Posted 2006-12-20 Print this article Print

OK, well, Tim Berners-Lee didn't create the Web so he could measure traffic, or count page views, or deliver top search terms and the number of blogs created per second. But in the age of Bubble 2.0, those measurements are certainly the focus of concern. Only problem is, they're all a bunch of bunk.

Or not. Depends on who you ask. And that's the problem--and the joy!--of working on the Web. Why am I even complaining about this? Because this week Google released its top searches for 2006. Only problem is, the terms they released aren't the top searches. More likely, they've been scrubbed of salacious and competitive material. Moreover, there's no overlap between the top search terms on Google, Yahoo and AOL.

No overlap? Are you serious? I'm more inclined to believe that the results are just marketing ploys--the engines are aiming at different markets, and they want to reinforce that brand image.

The unreliability of top search results is just the latest in a long string of unreliable data that we consume and bloviate about on a regular basis:

  • Traffic stats are unreliable -- How many disagreements about traffic stats have there been this year? Digg disagreed with ComScore's numbers., too, said ComScore's numbers were way off. And then there's the recurring problem of Alexa, which nobody seems to trust. Despite the disagreements we muddle through and we reach valuations and append dollar signs. But that just goes to show that traffic stats, like money, are just a kind of poetry.

  • Click fraud stats are unreliable -- Lordy, didn't we have fun hashing this one out this year. Do you believe Google when they say click fraud is 2 percent? Or are you more inclined to believe it's 10-15 percent? It doesn't really matter, one AdWords buyer told me this week, because as long as click fraud is the same for all our advertisers, nobody can tell the difference.

  • The number of blogs is unreliable -- When's the last time you take Dave Sifry's stats without a grain of salt? Sure, it's compelling to know that a new blog is created every two seconds, and that the blogosphere doubles in size every six months. But any true picture of the size of the blogosphere, not to mention the volume of posting, is more complicated.

  • The number of people actually actually using social networks is unreliable -- Quote unique visitors and page views all you want. Nobody knows how many people actively continue to use social networks. Even the big media pieces on MySpace's attrition rate have to rely on conjecture.

Now I'm not saying all these phenomena aren't big, big, big ideas, or cash cows, or cultural milestones. I'm just saying I'm tired of all these unreliable statistics. I'm in existential quandary here. It's enough to make me want to cut off my computer and actually go home for Christmas. |

Submit a Comment

Loading Comments...
Manage your Newsletters: Login   Register My Newsletters

Rocket Fuel