Profiting from Non-existent Domains

Over at Domainincite, Kevin Murphy has put forth this observation: 

"For the first time, registrars will be able apply for and run new top-level domains, giving them unprecedented insight into registry-level data.
If they also act as registries, registrars will, for example, be able to see what non-existent domains in their TLD get the most type-in traffic."

Kevin's comment led me to wonder about just how profitable non-existent domains (NXD) might be to a registrar... so I decided to gather up some empirical data.

In May 2009, the registrar Dynadot put out this marketing pitch: 

"VeriSign, the central registry for COM domains, recently released a new tool called Data Analyzer. Data Analyzer allows you to check the amount of traffic a domain name gets, even if the domain does not exist."

As this tool would clearly be of interest to the domain tasting crowd, it made sense to check out Dynadot's registration numbers over the course of the last year.  In one year they went from 229,000 to 282,000 registrations (in .com) -- not a bad growth rate (especially in a lousy economy). 

Registry contracts already stipulate: "Nothing in this Agreement shall preclude Registry Operator from making commercial use of, or collecting, traffic data regarding domain names or non-existent domain names",
and registrars are already buying the data.

The new ICANN cross-ownership policies won't really affect this NXD market... registries will simply continue to sell their data to registrars as there are no contractually-based prohibitions in place to deter such practice.