Jim,

This really is a misleading title, even if you are only talking about traffic.

David

For the sake of discussion on this, I'll share our real unique visitor numbers vs. what Compete reports. We had over 200,000 unique visitors to our www.uwcu.org Web site in September 2007. However, Compete says we had under 40,000. My numbers are actual since they do not use a "unique methodology created by experts in the fields of mathematics, statistics and the data sciences to aggregate, transform, enhance and normalize data in order to estimate U.S. Internet traffic"--it just is an actual count (with the appropriate filters to remove non-human traffic).

Perhaps then, Compete is better used to measure trends rather than talk actual visitor numbers. It does properly reflect the relatively flat growth we've seen in unique visitors over the last year. However, I wonder if the spike for Mint.com is more of a result of the math involved in estimating than actual numbers due to the buzz Mint has created recently causing a huge (but potentially short-lived) surge in traffic.

@David,
The title wasn't meant to be misleading...just an observation from the Compete dataset.

@Eric
Thanks for sharing your traffic numbers. Good point...third-parties such as Compete, comScore, Nielsen/NetRatings are providing estimated numbers based on click-stream data. Their estimates can be very wrong...so it is risky to read too much into any specific number, especially where their is an unusual amount of new activity, such as Mint experienced.

The point I was trying to make was that Mint had a significant spike in traffic. The comparison to a traditional bank, was just to give the number some context.