Unfortunately We Were Right Regarding Flux

Yesterday we wrote about the need for a real Web comparison engine. In the article, we discussed Flux and the apparent traffic growth they were seeing via a chart that Mike Arrington posted earlier this week. The chart (now removed) showed a huge spike in traffic which Arrington believed came from Viacom's cash bankroll. I believed the spike actually came from the widget setup that Flux was using (now also removed on Techcrunch).

Unfortunately it appears we were correct. Arrington has posted an update to his article:

Flux may be taking off, but not as much as we thought. The January spike in the comScore data was a little suspicious, so we asked comScore to double-check that it was right. Turns out it was an “artificial increase” due to traffic not requested by users. comScore is revising its numbers to filter out the bad data. We’ll let you know what those are as soon as we hear.

I said "unfortunately" above because when the numbers on a tracking service are off, we all lose. comScore seems to be the service that many analysts and bloggers trust more than the others (Alexa, Compete and Quantcast).

Side note, I do wonder if Compete is the big winner in this mess as their numbers are much lower for Flux than comScore's initial numbers. I'd love the opportunity to speak with someone at comScore once the new number for Flux is revealed. Clearly they have some explaining to do.

At the end of the day analytics issues impact advertisers and publishers. Advertisers believe the charts and graphs shown to them without doing any further research many times. Does this mean that a site like Flux may have been able to pick up ad buys when they should have went to the competition (KickApps, Ning, Magnify, etc.)? This is the bottom-line fear when the numbers are wrong. My advice is simple, never trust the numbers alone -- no matter the source.