Well, two years later, that trend continues. While we have as many U.S. users as ever, on a strict percentage basis, they are declining as the Alexa Toolbar grows in popularity across the globe.
I thought I'd update the chart so you can see for yourself.
| Country | Users | (previously) |
|---|---|---|
| China | 16.44% | (18.46%) |
| United States | 14.28% | (36.91%) |
| Brazil | 3.82% | N/A |
| Japan | 3.64% | (3.80%) |
| United Kingdom | 3.11% | (4.49%) |
| Taiwan | 2.91% | (1.72%) |
| Hong Kong | 2.55% | (4.59%) |
I dug around the web a bit looking for some stats about the estimated actual distribution of web users across the globe and found this site. They break it down by region, and estimate that North America now accounts for only 21.2% of Internet users worldwide. I don't know if they are right, but when I add up Alexa's toolbars from the United States, Canada and Mexico I get 18%. Asia is apparently supposed to account for 35.6% of Internet users worldwide. Again, it is hard to know if their numbers are correct, but after adding up the Alexa Toolbar users from top 10 Asian countries I was at 30%.
Historically Alexa has avoided doing any "normalization" of our data. What this meant was that a toolbar user was a toolbar user and was counted in our stats. We didn't pick and choose users. But, as the Internet has grown into a global phenomenon we've had to make some adjustments from time to time. Now we are actually very picky. In fact, the vast majority of toolbar users are not counted at all. Ironically, by filtering out a large number of Alexa Toolbar users from our panel the data is actually improved. The Alexa panel more closely matches a typical distribution of Internet users worldwide.
Still, even with normalization, this global distribution has some interesting side-effects. I was contacted about a month ago by a colleague who had noticed that Alexa had shown a steady decline of his site's reach over the course of the year, which stood in stark contrast to his own internal stats which showed a steady increase in users. The answer to the apparent discrepancy is that Alexa's Reach number is not the same thing as visitors. The reach is counting the percent of Internet users worldwide who visit a site. In the case of my colleague's site, his global reach was declining, not because he was losing visitors, but because the rest of the Web was growing so quickly. As people are coming online in China, Korea, Vietnam, Brazil and countless other countries, they are counted as users in Alexa's panel, but they aren't visiting his site. So, in essence, his percent of global traffic is declining, even as his traffic goes up.
We will continue to investigate ways to improve the Alexa panel. I personally would like to find a solution that would provide a better measure of Visitors that doesn't suffer from the issues that I see with Reach. Having country panels might be a good start.
