|
Verisign, SiteFinder,
and WLS |
|
|
Note: At the time this article was written, two "services" either proposed
or implemented by Verisign were under consideration by the Internet community and/or
ICANN. Subsequent to this writing, both proposals were delayed, shelved or for some reason
not continued or implemented, and the Internet has remained somewhat untouched by arbitrary
actions on the scale of those discussed here. We should remain diligent however, because
the same players are still in the game, and the motivations of those players are unlikely
to have changed much.
Although the following comments would probably apply to all gtld registries equally, because Verisign has instigated the discussion and because they have provided us with two examples of services which propose to add value to non-performance, the comments will refer to Verisign and their two proposals - Sitefinder and WLS - instead of generic principles. The Value of Non-Performance As a registry operator, operating under contract with ICANN, Verisign's role is to administer the registry or the database containing all valid registrations of .com and .net names. Their job, simply put, is to ensure that a name that is appropriately registered is maintained in the database. WLS and SiteFinder both are designed in such a way that Verisign will maximize their cash intake precisely by not doing their job, or at least by not doing it as well as they could. Here's how.... With WLS, Verisign's proposal is to limit subscriptions on any particular domain name to the first purchaser, or one subscription per name at a time. Once that subscription is sold, it must either expire or be used before another one may be sold on that name. Such limiting means that Verisign must sell subscriptions on as many different names as possible to enjoy maximum revenue. Another limiting factor is that many names are currently under multi-year registration, and will not expire within the one year lifetime of a WLS subscription, making those names extremely unattractive targets of WLS subscriptions. One of the ways they propose counter that limitation is to promote the WLS as a benefit to consumers by describing it as a sort of insurance policy in the event of inadvertent removal of their domain name from the registry. This "insurance" would also be of purported benefit in cases where registrants neglected to renew their domain names for some reason. Since it is much cheaper for the registrant to renew than to buy the "insurance", it will be hard to sell subscriptions on that basis. (It would make more sense to take the money required for the WLS subscription and apply it to renewals). If, however, Verisign could "accidentally" drop the name from the registry and thus "use up" the current subscription on it, they would be able to then immediately sell another subscription on the name - AND - they would also demonstrate the "value" of having such a subscription to existing domain owners - on all names, not just those that will expire within the next year. The more Verisign can get away with not maintaining valid names in the registry as their contract stipulates, or more subtly, the more they can promote the fear in consumers that the registry may not do their job correctly, the more money they can make. Since Verisign would obviously get in trouble with consumers and/or ICANN if they accidentally dropped too many names, doing so would not be in their best interest, but if they can get away with dropping an "acceptable" number of names, they will benefit from being able to sell new subscriptions on those names (increase the churn rate), and also from validating their sales pitch for the service on all names they haven't dropped. The goal will be to find that "magic level" of incompetence which will maximize their profits. Because of the relatively high profit potential to the registry on WLS subscriptions as opposed to normal registrations, a seemingly small percentage of decline in performance under the ICANN contract could yield the registry operator a significant boost in income levels, making it highly tempting for those making the decisions at Verisign to actively pursue controlled non-performance. |
|
|
SiteFinder stinks too.. The scenario for SiteFinder is very similar. Since Verisign will only be able to monetize traffic on names that are not in the registry database or (as the service was originally implemented) that do not have valid name servers in the database, SiteFinder will make Verisign more money by having fewer names actually registered. Since pay-per-click rates on advertising are relatively low, Verisign itself would probably benefit more from the actual registration of a name than from miscellaneous traffic on that name if it were unregistered. If a way were found to raise the profitability on the traffic to those names (or some of them) beyond what they make for selling the registration of that name, this conflict would then become more pronounced. Verisign could have their cake and eat it too though. Any name which is registered and drawing significant traffic would make Verisign more money in a 24 hour period if they could capture that traffic, and all that would be required would be to inadvertently lose the name servers for a that name for a while. They still get paid for the registration, but they also get paid for not getting it in or maintaining it in the database promptly or correctly. Or.. in the case of a name which has been put on Registrar-Hold for any reason, if and when that Registrar-Hold is removed, Verisign will benefit more the longer they can delay the actual re-activation of the name. This will be especially true for names which already draw high traffic levels. SiteFinder is a bad idea for a lot of reasons, but this is an example of why it is bad because its goal is to make Verisign money. |
|
|
Conclusion These are just a few methods, and I am sure there are several ways I haven't thought of whereby the more poorly Verisign performs under their ICANN contract, the more they will make, and again, the goal for them will be to find that level of sub-optimum performance which will maximize profitability while still performing at "acceptable" levels. This is an inherent problem with any system where one must weigh cost against the benefit of perfecting performance, but as you increase the value of non-performance, the scales will tip more in that direction. For an organization such as ICANN, who have stated that part of their mission is to maximize the performance of the DNS system, does it make any sense at all to employ contractors who are proposing ways in which they can get paid more by offering less than perfect performance?
Copyright 2003 John Vogel
|