Why Advertisers Pay for Spots on Bogus Web Sites

The front page of newyorktimes.cm. Click to enlarge.

There is a fascinating, if dated, CNN Money story from a few years ago about a man who owns 300,000 domain names. Most interestingly:

And what few people know is that he's also the man behind the domain world's latest scheme: profiting from traffic generated by the millions of people who mistakenly type ".cm" instead of ".com" at the end of a domain name.

Try it with almost any name you can think of -- Beer.cm, Newyorktimes.cm, even Anyname.cm -- and you'll land on a page called Agoga.com, a site filled with ads served up by Yahoo (Charts,Fortune 500).

Ham makes money every time someone clicks on an ad -- as does his partner in this venture, the West African country of Cameroon. Why Cameroon? It has the unforeseen good fortune of owning .cm as its country code -- just as Germany runs all names that end with .de.

The difference is that hardly any .cm names are registered, and the letters are just one keyboard slip away from .com, the mother lode of all domains. Ham landed connections to the Cameroon government and flew in his people to reroute the traffic. And if he gets his way, Colombia (.co), Oman (.om), Niger (.ne), and Ethiopia (.et) will be his as well.

It looks like the domain names have since switched hands, and Agoga.com seems to have vanished (see an archived version here). The interesting economic question here is, how can such sites be profitable? Why do Yahoo and Google, and by extension their advertisers, help subsidize these deceptive practices and thus offer some sort of tacit approval?

Imagine two ways (there are countless more) for an Internet user to come across an ad for a particular kind of car:

• Finding a blog or news article about the Toyota Prius and clicking on a contextual link along the side of the page.
• Trying to type in "prius.com," getting taken to an advertisement-laden page because of a typo, and seeing the same ad and clicking on it.

A blog post takes at least some effort to produce and is much more likely to be informative to the audience. A typosquatting site, on the other hand, can be produced almost automatically and costs about $7 a year for the Web hosting service. The former looks to build reader relationships over time through continued quality content, while the latter deceptively attracts one-off hits from poor spellers and hasty typists.

However, clickthroughs on the typosquatting sites are probably more valuable for advertisers. Not everyone who reads a story about a Prius is interested in buying one, but someone who bothers to type in (something close to) "prius.com" certainly may be. If advertisers take a moral high ground and refuse to advertise on such sites, they are probably costing themselves money.

I imagine that the typosquatting domain name business is becoming less profitable these days. Although old habits die hard (many people still use Internet Explorer 6, despite many calls for its abandonment), most modern Internet browsers have a built-in search bar at the top, in addition to the address bar. Google's Chrome browser has only one bar at the top for both URLs and Google searches, so a user typing in "shoes" is going to see a list of shoe-related searches populate as he types and is less likely to type "shoes.com" (or some misspelling thereof). When people look for such products via a search engine, the bogus Web sites mentioned above will be ranked very lowly and thus won't generate many clicks.

In addition, owners of such sites are often sued by the legitimate owners of the respective copyrights or products and thus forced to surrender the domain names (though no one can sue for the rights to a bogus site whose URL is based on the word "shoes").

Of course, these sites can make money in other ways besides advertising, such as installing malware on visitors' computers in attempt to gain access to financial information.

Thanks to Matt McFarland for the tip.