Which domain extensions do people prefer? How do these preferences vary by country? This article reports the results of a poll of domain extension preference by country of residence conducted from May through October, 2003 by Domain Name Journal and the DomainState forum. Approximately 133 people from 34 different countries participated. This poll has limitations that we will discuss, but it does measure the extension preferences of domain registrants and developers in a manner that has not been done elsewhere and it produces some interesting results.

The main page of the poll linked to pages for individual countries to measure extension preference by country. The poll asked the following question—“Which extensions are worthwhile for registering and/or developing domains—considering the quality of domains available and the prices to obtain those domains?” Respondents could select one or more of the following extensions: .COM, .NET, .ORG, .INFO, .BIZ, .TV, .CC, .WS and the country code extension for their country of residence. We originally included .US and .CN extensions regardless of participant country, but decided not to include them in the analysis to avoid murky issues of registration of country extensions by non-country-residents.

Limitations

The first limitation of this poll concerns question wording. We could have asked a question that does not address domain availability or price, such as—“Which extensions are worthwhile for registering and/or developing?” However, using a car analogy, this would be like asking—“Which car would you prefer? Mercedes Benz or Honda?”—without mentioning the price difference. People would probably respond—“If you are giving a car away, I’ll take the Benz”—but this response, ignoring price, does not shed much light on real world markets because price matters in the real world. A more meaningful question is—“Which car do you prefer given its price? Mercedes Benz or Honda?” Answers to a question that includes price give more insight into real world purchases. For this reason, we explicitly included price and availability in our question. A simple question that ignores real differences in price is not ideal, but a question that considers differences in price and availability is not ideal either due to its complexity. The difficulty in balancing these concerns in question wording is one limitation of this poll.

The second limitation of the poll concerns possible bias in the population sampled. Members of a domain forum, mainly domain investors and developers, probably do not represent the general population or even all domain registrants. They are likely to be much more informed about new extensions and willing to take risks on new extensions. If one wants a representative sample of the general population, then this population and the results of this poll are probably biased toward new domain extensions. On the other hand, one could view a population with above-average expertise concerning valuation of a scarce commodity as a preferable population for evaluating future value of that commodity. For example, if one wanted to assess the future value of mineral or transmission bandwidth rights, perhaps one would prefer a survey of specialized industry experts over a survey of the general population? From that perspective, perhaps the knowledgeable population surveyed could be viewed as a strength of the poll rather than a limitation?

The third limitation concerns sample size. This poll would be better and the results more powerful with 1,330 respondents instead of 133. Individual results for countries with only a few respondents are not very accurate. However, the results of 133 people from 34 countries are probably better in aggregate than anecdotal evidence or no data at all.

Results

Detailed poll results are shown in the following table. The numeric columns show the percentage of votes for each extension by respondents for each country. “Country Code” is the domain extension for a respondent’s country of residence (such as .US for United States, .CA for Canada, etc). The last column shows the number of respondents for the country (estimated by the maximum number of votes for any extension within that country). Countries are listed in descending order of number of respondents. For example, of the 44 respondents from the United States, 77% selected their country code (.US) as being worthwhile and 100% selected .COM as being worthwhile.

                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                     

Table: Percentage of Country Respondents Selecting Each Extension with Countries Sorted by Number (N=) of Respondents
Country:	Country Code	COM	NET	ORG	INFO	BIZ	TV	CC	WS	N=
United States	77%	100%	61%	32%	50%	32%	0%	0%	2%	44
Canada	100%	94%	29%	29%	24%	6%	0%	0%	0%	17
United Kingdom	100%	73%	45%	27%	64%	45%	0%	0%	0%	11
Germany	100%	89%	44%	33%	89%	33%	0%	0%	0%	9
Australia	100%	100%	100%	25%	25%	25%	0%	0%	0%	4
Netherlands	100%	100%	100%	75%	75%	25%	25%	0%	25%	4
Norway	75%	100%	50%	25%	75%	0%	25%	0%	0%	4
China	100%	100%	50%	0%	50%	50%	0%	0%	0%	2
Croatia	100%	100%	100%	0%	50%	50%	0%	0%	0%	2
Denmark	50%	50%	50%	50%	100%	50%	0%	0%	0%	2
France	50%	50%	50%	0%	100%	50%	50%	0%	0%	2
Greece	50%	100%	50%	50%	50%	0%	0%	0%	0%	2
India	50%	100%	50%	0%	100%	100%	0%	0%	0%	2
Ireland	100%	0%	0%	0%	0%	0%	0%	0%	0%	2
South Korea	0%	50%	0%	50%	50%	100%	0%	0%	0%	2
Malaysia	0%	100%	50%	0%	50%	0%	50%	0%	0%	2
Russia	100%	100%	0%	0%	50%	0%	0%	0%	0%	2
Singapore	50%	100%	50%	50%	100%	50%	0%	0%	0%	2
Taiwan	100%	100%	0%	50%	50%	0%	50%	0%	50%	2
Turkey	0%	100%	0%	50%	0%	0%	0%	0%	0%	2
Belgium	100%	100%	0%	100%	100%	100%	0%	0%	0%	1
Finland	100%	100%	100%	0%	100%	0%	0%	0%	0%	1
Iceland	0%	0%	0%	0%	100%	0%	0%	0%	0%	1
Italy	100%	0%	0%	0%	0%	0%	0%	0%	0%	1
Japan	100%	100%	100%	100%	0%	0%	0%	0%	0%	1
Mexico	100%	0%	0%	0%	0%	0%	0%	0%	0%	1
New Zealand	0%	100%	0%	0%	0%	0%	0%	0%	0%	1
Philippines	100%	0%	0%	0%	0%	0%	0%	0%	0%	1
Poland	100%	0%	0%	0%	0%	0%	0%	0%	0%	1
Portugal	100%	100%	100%	0%	0%	0%	0%	0%	0%	1
Saudi Arabia	0%	100%	0%	0%	0%	0%	0%	0%	0%	1
Sweden	0%	100%	0%	0%	0%	0%	0%	0%	0%	1
Switzerland	100%	100%	0%	0%	100%	0%	0%	0%	0%	1
Thailand	100%	100%	0%	0%	0%	0%	0%	0%	0%	1
AVERAGE:	71%	77%	35%	22%	46%	21%	6%	0%	2%	4
WEIGHTED AVE:	80%	89%	47%	29%	50%	26%	4%	0%	2%	133

The “AVERAGE” row (second to last) shows the average percentage of votes for each extension across all countries, where each country has equal weight. The last row, “WEIGHTED AVE,” shows the percentage of votes for each extension for all 133 respondents, effectively weighting each country by the number of country respondents.

Based on the “WEIGHTED AVE” row, the main result of the study is four overall tiers of domain extension worth. The first tier includes .COM (89%) and country code extension within country (80%). The second tier is .INFO (50%) and .NET (47%). The third tier is .ORG (29%) and .BIZ (26%). The fourth tier is .TV (4%) and .WS (2%). A last extension, .CC, received no votes. Having said this about tiers overall, there is some variation across countries.

A second result is the relative popularity of .INFO (89% and equal to .COM) in Germany and several other European countries compared to the global weighted average for .INFO (50%). This is consistent with anecdotal accounts of high .INFO sales prices and registrations in Germany.

A third result is the relative dominance of .CA (100%) and .COM (94%) in Canada. Among the countries with many respondents, Canada seems to be the most conservative with respect to breadth of extensions. In other words, if you try to sell an .INFO or .BIZ to Canadians, they are likely to say—“Eh?” At the other end of the spectrum, people in the Netherlands, Norway, and Denmark seem to value a wide breadth of extensions.

Conclusion

This poll asked domain discussion forum members how worthwhile extensions are for domain registration and/or development—considering the quality of domains available and the prices needed to obtain those domains. In this context, “COM is King” still holds true overall, but country code extensions are often equivalent within their respective countries. Also in this context, among the new extensions introduced by ICANN during the past couple years, .INFO has moved up next to .NET and .BIZ has moved up next to .ORG. 

There is considerable variation among counties in the relative evaluation of domains. Some countries are relatively conservative in their focus on .COM and country-code extension, while other countries are more receptive to a wider breadth of extensions. As with stock analysts’ projections of earnings per share, there are no guarantees concerning the future. This poll measures people’s subjective judgments. Any of these extensions could succeed and any could wither away. However, the combined wisdom of 133 people from 34 different countries is probably a pretty decent window into the future.