ICANN has started its historic and controversial program to expand the number of generic Top-Level Domains (gTLDs). This essay outlines the factors needed for the program to create economic value, warns against a cognitive trap that complicates selection of a new gTLD and considers the value contribution of the registries. I will not go into relevant macro measures, but I examine the problems associated with the popular measure of simply counting the number of registrations.

The key to understanding the program’s economic impact is to follow the theories of economist Paul Romer and look at how the rearrangement of resources creates value. ICANN’s program increases the supply of resources that registries have for creating value. Value creation by registries can come from: (1) introducing new TLD signals for things like location, community, and social responsibility (for example, .nyc for New York City, .music to signal community, and .green to signal environmental corporate responsibility); (2) combining information, such as in the .tel model, which provides contact information for the companies using the gTLD; and (3) introducing a gTLD that competes with .com.

Given the new resources provided by ICANN, the burden now lies on the registries to innovate. But they have to be careful of cognitive biases in choosing among the gTLDs. For example, a registry that chooses the proposed .music should ask itself, “Is there value in .music?” The temptation is to ask the far easier “Do we love music?” Not the same thing, but studies show that we often answer an easier question instead of a harder and more relevant one, and that we’ll do so without noticing the swap. (For details on cognitive error traps, see Daniel Kahneman, Thinking, Fast and Slow. I have warned against cognitive biases in gTLD value estimation and in domain name appraisals.) Another trap is reliance on the popularity of key words in social media, an approach that flopped with the recent failure to predict the success of presidential candidates.

Remember, there is no easy way to measure new gTLD value creation. The domain name industry has focused on registrations, but that’s because they are easily measured and the information is publicly available. Number of registrations does provide a viable measure of a registry’s profits, but the registrations may be defensive by brand owners rather than value creating. (For a discussion of alternative measures, see “The Economics of Well-Being” by Justin Fox, HBR January-February 2012.)

New signals and combinations of information, á la .tel, can be value adding for established companies as well as new ones. But switching costs will probably keep most com-branded companies from making the jump. So new companies may converge on a new gTLD that competes with .com while existing companies will more than likely register their brands under a large number of the new gTLDs as a defensive measure. Put all the registrations together and there will be enough revenues for the com-alternative gTLD to be viable.

One reason for gravitating to a com-alternative gTLD is that new companies might feel constrained by the unavailability of desired .com names and thus have a motive to find reasonable alternatives. (See Why Dominant Companies Are Vulnerable by Kyle B. Murray and Gerald Häubl, Sloan Management Review December 2011.) This is especially true because emerging brand owners don’t have to acquire any new skills in order to adopt a new gTLD.