ICANN recently commissioned a report from a domain auction company to see whether it would be a good idea to auction Top-Level Domains (TLDs) that have multiple applicants. Remarkably, the domain auctioneers came to the conclusion that auctions are a great idea, which they surely are for some people. But are they a good idea for ICANN? And if ICANN admits they can’t evaluate competing applications on their merits, how can they keep the process from turning into another speculative land grab?

On the one hand, there’s no question that ICANN wasn’t able to run the beauty contests for the last two rounds of TLD applications successfully, since they took forever and produced arbitrarily results. But auctions are virtually guaranteed to leave all the high profile domains parked with domain speculators.

The experience of .MOBI is instructive here. (I am on the .MOBI policy advisory board, as the ICANN ALAC designee.) They’ve tried a bunch of different approaches to distributing domain names. They auctioned off some “premium” names, and they’ve done an RFP for others, somewhat like the beauty contest except that the registry picked the names, not the applicants.

The auctions were a financial success, but a failure for domain users. Most auctioned names are just parked with ad sites, a lot of which don’t even work properly on mobile phones, so we can see that speculators routinely outbid any real applicants who had plans for the names. See, for example, flowers.mobi which cost $200K, fun.mobi ($100K), hot.mobi ($40K), and stockquotes.mobi ($27), all of which are mobile-hostile parking sites. Given the pervasive culture of speculation in the domain registration business, I see no reason to expect ICANN auctions to produce any other result.

There haven’t been many .MOBI RFP’s, but the few they’ve done seem to have worked, most notably weather.mobi which is run by the Weather Channel and provides actual weather information. I don’t know the details of their agreement with mTLD, but my impression is that rather than charging a lot, .MOBI got a contractual promise to build a useful weather site. I do think a key reason this can work is that they picked names and topics whose meaning is straightforward, making it a lot easier to evaluate proposals, and I’d expect the same process could work for TLDs.

With respect to all the money that auctions can bring in, .MOBI has spent a lot on tools for building mobile web sites, e.g., the device atlas at deviceatlas.com which lists the characteristics of all the browsers on all the models of phones available so that sites can customize themselves to the device on which they’re displayed. (Mobile web sites are a lot like PC web sites a decade ago.) That’s fine for .MOBI, where mobile web sites are a major part of what they are, but I don’t see anything analogous for ICANN.

In particular, I have grave reservations about the suggestion that ICANN should engage in yet more mission creep with an even larger budget and use the auction proceeds to fund virtuous causes. Given ICANN’s great difficulty in managing itself, and its inability to deal consistently with all of the virtuous TLD applications since 2000 (how come they rejected .MAIL but approved .ASIA?) I cannot see any reason to expect other than yet another bureaucratic mess. If ICANN gets lots of extra revenue from auctions, they should remember what business they’re in and use it to reduce the 25 cent domain tax instead.

I don’t have a good answer for dealing with competing applications. In some cases, where all the applicants have plainly venal motives, there’s no question that there’ll be an auction, the only question is who keeps the money, so it might as well be ICANN. But if you have a rich speculator vs. a non-profit or startup that has a more or less plausible plan for the domain, you’re back to beauty contests. So I don’t necessarily think it as a bad thing that, as I noted last month, it’s unlikely that this process is going to go anywhere soon.