Why did losers not value the domain name as high as the winner?
Let me start by saying I’m not a big fan of the idea of community priority and similar mechanisms for determining which applicant gets a particular top level domain name.
But I’ve also been thinking lately about the auction process and its effect on the long term health of new top level domain names.
When an auction is used to resolve a new top level domain name contention set, the company with the highest perceived value of the domain name gets it.
That makes sense. But does it mean the other applicants were all wrong in their calculations of how much the TLD was worth?
Consider a five-way auction for .something. Presumably, every applicant ran calculations ahead of the auction to figure out how much it was worth. The auction ends with just one of them victorious, meaning that only one of the applicants calculated the value to be greater than the closing price.
Did the other four applicants make an error in their calculations? Or did the single firm that won make an error?
Out of the five, only one had projections high enough to justify the price. That’s a bit scarey for the future of the top level domain name at issue.
Yes, I understand a lot more goes into it. A TLD could be worth more to a company that has a direct connection to a related industry. Some applicants have greater efficiencies and channels to market. Some applicants have a lot more money to play with and can take greater risks.
At the same time, if I won a five-way race, I’d have reason to doubt my forecast.
George Kirikos says
That’s called the “Winner’s Curse” in auction theory.
Philip says
It just comes down to partys having limited resources and probably a wider agenda for some with regards to other TLDs they may also wish to secure, within their financial constraints.
Philip says
They may actually think a specific TLD is worth the price paid but if paying that price undermines a wider buying strategy, the bigger picture for such acquisitions needs to be taken into consideration.
Monte Cahn says
Andrew – you raise a good point. Remember though that the highest bidder in these auctions pays the 2nd highest bid price, so actually 2 of the 5 in your scenario felt it was worth at least that price.
Also we all know that in auction theory, having an ascending step type clock auction generally results in the highest winning bid where a single sealed bid auction generally results in the lowest winning bid, simply due to the fact that all the participants cant see who is still in each round and at what step price. Sealed bid participants just place a single sealed bid at their budget and never know what price any of the other participants are bidding so it never influences their decision to place another bid or to keep bidding above what they can afford.
It is a fact that many of the winning bidders in the current TLD auctions have over bid their original budgets to win the auction, some well over their budget which could put financial stress on their overall business plans even if they are successful in raising additional funding to make up the difference.
Just some points to consider.
Constantine Roussos (.MUSIC) says
Andrew,
The bottom line is the ones with deeper pockets will always win if all conditions are held equal. In the case of restricted community TLDs winning auctions, the path to success is more difficult because their restrictive policies that are custom tailored to serve their communities are at odds with maximizing registration volume to increase profits to offset significant auction losses.
I think this is more of an issue of public interest since each TLD is a case-by-case scenario. A .WEB TLD should be open and should be auctioned out to the highest bidder. However, a .HEALTH in my opinion should not because health is a regulated sector and should have certain enhanced safeguards to protect the public interest and not be launched under your typical “wild wild west” TLD launch approach.
The problem with the new gTLD Program is that it treats every TLD as being equal. Truth is that users will expect certain safeguards for a .MED or .PHARMACY while they would not expect the same safeguards for a .ONLINE unless someone can convince me otherwise. GAC and ICANN have at least agreed about this issue but yet the issue of auctions still looms large.
One the issue of communities, the fear of gaming has created a flawed CPE process that is unclear, inconsistent and unpredictable. We all expected CPE to be an objective exercise as mentioned in the AGB.
While one can argue that a community TLD is worth more to a community applicant the math does not add up. Connections to the community that does not translate well in regards to the ROI or financial projections and flexibility of altering its business model. A community applicant — with appropriate policies — has higher compliance costs due to restrictions (e.g. we have eligibility criteria, 2-step authentication, name selection, globally protected marks list, music content/use policies, registrants appealing to fix incompliance pertaining to IP infringement etc).
So when you couple the appropriate safeguards with less registrations than an open model and lower ability to maximize revenues (which is not the primary objective of community applicants) then under what economic reasoning should a community applicant engage in an auction? This problem is further expanded when you cannot possibly compete with a Google, an Amazon, or a venture-backed portfolio applicant since they have deeper pockets, economies of scale, open policies and diversification tactics to recover income if they overpay (e.g. many private auctions and domain proceeds from other TLDs).
I am pleased that many in the ICANN community are having second thoughts about auctions since each TLD is a case-by-case scenario and the public interest can be at stake for some sensitive TLDs on the GAC sensitive regulated string list.
Recent compliance discussions over illegal drugs and ICANN’s role when it comes to serving the public ((http://online.wsj.com/articles/icann-regulators-clash-over-illegal-internet-drug-sales-1414463403 and https://www.techdirt.com/articles/20141029/07223828977/fda-is-angry-that-icann-wont-just-censor-websites-its-say-so.shtml) has shed some positive light on the issue of selecting the right registry for the right string with the right safeguards and appropriate compliance needed for that particular string.
Public auctions do pose a bit of a conflict of interest for ICANN when it comes to CPE though. When a CPE applicant prevails then they never recover their $22k for the evaluation and are subject to a public auction benefiting ICANN’s bottom line. If an applicant wins CPE then ICANN has to pay the $22k CPE fee and lose out on a highly-valued public auction with proceeds ending in ICANN’s pockets. However, in ICANN’s defense they have been very careful with this conflict thus far which is commendable.
Many in the community are discussing the conflict of interest reputation of ICANN when it comes to the power of companies such as Google within the ICANN framework. For example, how are Google top executives permitted to be placed in positions of possible conflicting influence with ICANN decisions with respect to ICANN decisions and contention set influence. For example, Google’s VP and ex-chairman of ICANN, Vint Cerf, was selected by ICANN to be on its 5-year strategy panel (https://www.icann.org/news/announcement-2013-07-15-en). How is that not a conflict? Such a decision should have been made after all contention sets with Google were resolved. Also Google executive chairman Eric Schmidt is on the Board of Directors of the Economist Group (http://www.economistgroup.com/results_and_governance/board.html), whose Economist Intelligence Unit is responsible for grading CPE. How is that not a conflict? ICANN should not have selected the Economist as an anonymous panel when Google’s chairman has direct influence as a director within the Economist. If Google is in your contention set then your odds are further stacked up against you if Google wants a string. Not only is Google beating you with deeper pockets in auctions for strings they want, they are beating you with influence at every level.
Even so, call me an optimist but I still have faith and believe ICANN will do the right thing. Hope is all we have 🙂
Colin says
In .CLUBs case we were ecstatic. Our analysis included quantitave and qualitative measurements.
Andrew Allemann says
Your case is a good one to point out that different business models yield different valuations. You guys are going all in on one string vs. a portfolio play.
I think you also got a relatively good deal compared to some of the current prices 🙂
Phil Buckingham says
Andrew, a very interesting point . Colin is right – a mixture of quantitative and quantative factors , but clearly every applicant auction strategy is different . Guess it all boils down to how badly applicants want their TLD and whether these new TLDs are perceived as a game changer or not. Forecasts are forecasts . They are certainly very very different to back in 2011 /2012 when done for application. so what price/ factors for.hot – amazon , donuts , b johnson confirmed ICANN auction 19 November . just a thought – Do a poll on your readership .
bladel says
It’s an interesting point to ponder, but setting aside risk tolerance and operational efficiency, I see it as a measure of the Applicant’s confidence in their demand projections, or ability to generate demand.
Alex Tajirian, DomainMart says
As George Kirikos correctly points out that the winner’s curve is very common. However, in general, assuming no winner’s curve, the highest bidder is the one that values the asset more than the rest, i.e., he/she benefits more from owning the asset than the other bidders. Yet, with new gTLDs why would someone in round 1 benefit more than the rest, given that they all have the same public information and access to the same resources? Hence, there is something wrong with allocation of new gTLDs through auctions, i.e., allocation myopia; ICANN should consider lotteries.