How much of the web is adult-oriented? ... 5.9%

Posted by Matthias Gelbmann on 26 May 2010 in

Other surveys

There are a number of surveys published on that subject. Their results range from 1% adult sites to 12%. Unfortunately, they don't give details on how the research was done, which makes it difficult to compare with our results. The presentation of the 12% survey gives the impression that the researchers were more drawn into the subject of study than into the statistical aspects of it. We have to admit: It's certainly more inventive than our little report.

What is an adult site anyway?

When calculating a percentage of adult websites, two seemingly trivial questions need to be addressed: what is adult-oriented, and what is a website.

What is considered adult-oriented depends of course very much on personal attitude and cultural standards. We try to match the mainstream judgment of the Western culture. That is not because we believe that it is the only valid yardstick, but because it happens to be our culture, and we feel comfortable making decisions based on that.

What is considered a website in our surveys might also be different to other statistics: we consider subdomains not to be separate websites. E.g. all the wordpress.com and blogspot.com subdomains are not counted. We also do not count redirected domains. E.g. live.com redirects to bing.com these days. It is not counted. Furthermore we restrict the statistics to the top 1 million websites, in order to limit the impact of domain spamming.

Why we make a distinction

We do not include adult sites in the technology surveys where we show popular sites and random examples of sites using a particular technology. Also, we don't include adult sites when we notify users which subscribed for technology change notifications. In the website information reports we do not show descriptions of adult sites and we do not insert an active link to the sites.

We make the distinction for two reasons. First, we do not want our visitors to stumble over adult sites, when they are looking for something completely different. And second, most advertisers don't want their ads to be shows in the vicinity of adult-oriented material.

By the way, we do include adult sites in the technology survey statistics, as we don't want content to have an impact on the number of technology usage.

How do we know which sites are adult-oriented?

Alexa includes an adult flag in their site info API, which we use. Unfortunately, we found that this is not sufficient, because the classification is wrong more often than we would accept. Therefore we do our own research by scanning the sites for any content that is typical for adult sites and re-classify them if necessary. Occasionally, we find sites that are classified incorrectly also by our own scanner. In that case, we try to adapt our algorithms so that it covers more and more cases. We also do manual classifications, but rarely and only if we cannot find a way to do it automatically. We did not have to change our algorithm for a while, which makes us confident that it is pretty good by now. Still, it is not perfect and we do not claim our figures to be 100% accurate.

And how successful are adult sites?

You probably know that some adult sites are in the top 100 most visited sites, but how successful are they on the whole? We have broken down the overall figure by rankings:

You can see that in the top 10.000 sites the share of adult sites is significantly higher, whereas in the top 1.000 it is lower again, but still more than in the top 1 million. That means, adult sites are doing pretty well, on average better than non-adult sites.

Nevertheless, we have no plans to change the scope of our own site any time soon.

P.S.: Some might think now: "So you have a list of some 60.000 popular adult sites. Could you, um, perhaps, well, actually publish it?" And the answer would be: no.

This entry is closed for comments.