Of course the only true way is a peek at the raw server logs.

Also if a site is aimed at web savy people who are more likely to have the toolbar installed the data will be skewed.

However, where Graywolf has jumped from 40k to 6k would indicate to me this does represent a substantial jump in traffic. But I can be sure of that as I have followed the blog over the same period of time.

For example, this site had an Alexa ranking of 200k a week ago and on Saturday it jumped to 8k, mainly because I wrote a post which attracted about 2,000 people mostly with toolbars installed.

Although even armed with that knowledge I still got a kick out of the spike in the stats.

Alexa reports back to base each time someone with it installed hits a site. Alexa then uses this data to extrapolate what their actual traffic is. So, if they get 100 hits on a day – they’ll assume that 1% of the visitors to that site probably had the toolbar installed and guess that they had 10,000 hits.

“Guess” is the operative word. Who visits SEO blogs? People interested in Search marketing and SEO. I would bet that the percentage of visitors with the Alexa toolbar or (Craig Raw’s Tools for FF) would be closer to 30-50%!

So Alexa, using the same math, assuming a 1% install base reports 100 hits as 10,000, when it actually could have been only 200 on an SEO site with a 50% Alexa toolbar install rate.

This is the dirty little SEO/Alexa secret, and it would make a good post in itself.