White Paper 17001

How many domain names are needed globally?

It is easy to find how many domain names are currently registered. An interesting question to ask is 'How many domain names are needed globally?'. While it is not possible to calculate a precise value, we can make an educated estimate. We provide that analysis in this paper. This gives us an indication of the growth prospects for the domain name industry, including what component of the market are likely to see the most growth. This analysis may also help inform whether additional new global top level domains (gTLDs) are needed.

How Many Businesses Globally

We start with the premise that the majority of domain names will be utilized by businesses. The number of companies registered on the major stock exchanges is only about 50,000 (you can see the statistics from the largest stock exchanges here). But most businesses are small and private, not registered on any stock exchange, so we must dig a little deeper for the relevant number of businesses.

For many countries information on how many businesses are registered (at different dates) is given in
this resource from the World Bank. For example, in 2007 (the last date with data provided) there were 2.5 million businesses registered in Canada. In 2005, the last year with data for the United States in the directory, there were 5.2 million businesses registered there, while the United Kingdom had 2.5 million businesses registered in 2007. The population of Canada in 2007 was about 33 million, the USA in 2005 about 300 million, and the UK in 2007 about 61 million. These data suggest a ratio of one registered business per 12 to 60 people.

It is likely the ratio will be less in developing countries, but the goal of this analysis is to predict the number of domain names needed in the medium term future, when development levels become more equal.

If we take the larger value, and assume it could be applied to the entire world population, this would suggest that, with a global population of about 7 billion people, the number of businesses globally might be about 120 million.

Alternative Method to Estimate Number of Businesses

We could estimate the number of businesses in the world another way. It might be argued that the number of businesses should scale as the Gross Domestic Product (GDP), rather than according to population. Data on the GDP of different economies is given here. Applying this method, the Canadian economy represents about 2.3% of the world total, so if we scale the 2.5 million registered Canadian businesses to obtain an estimate of the world total number of businesses, we get about 106 million businesses globally. If we use the USA figures of 5.2 million registered businesses in an economy that (by GDP) represents 24.7% of the global total, we scale to an estimate of about 21 million businesses globally.

These figures agree to an order of magnitude with those obtained in the previous section, although suggest a slightly lower value. Therefore, considering both techniques, let's revise our estimate of the number of businesses worldwide to about 100 million. The advantage of the second method is that it should more properly handle the mix of developed and developing economies.

How Many Domain Names Per Business?

While many businesses will use a single domain name, larger companies use additional domain names, both for different facets of the business (and/or specific product lines), and in multiple countries in which the business operates. I could not find good statistics on the average number of domain names per business, although a response on Quora estimates that a typical large technology company owns thousands of domain names, many of them unused. A lot has been written lately about domain names owned by the Trump organization, and it appears that the number is at least 3000. Many large companies now operate in a hundred different economies, or more, and need at least that many domains to protect their brand with country specific domain name.

On the other hand, the majority of businesses are essentially a single person, or a few people, engaged in one activity, such as a local contractor or an independent professional. Most of these need only a single domain name, or possibly two with the second used in marketing and directing to the primary company domain name.

It is probably safe to argue that the average number of domain names per business is higher than 1 and less than several thousand. I think the range is not really as wide as that though, and surely it is at least several, but on average, not more than several dozen. Let's somewhat arbitrarily assume that the average number of domain names 'needed' per business is 10.

Number of Business Related Domain Names

If we combine data from the previous two sections it suggests that the number of business domain names needed globally would be about 10 times 100 million, or about 1 billion domain names. It should be stressed that we have implicitly assumed that businesses in all regions have reached a technical, marketing and communication stage similar to that of major developed market companies.

What About Non-Business Domain Name Needs?

Of course many organizations that are not businesses register and need domain names - professional organizations, educational institutions, etc. It is perhaps even more challenging to estimate this number. One way to get an estimate is to consider the number of .org domain names relative to the number of .com. At the current date there are about 128 million .com domain names registered, and just over 10 million .org domain names. This might suggest that the needs of other organizations is of the order of 1/10 the number of domain names needed by businesses, although other organizations disproportionately use country code and alternative gTLDs, so it is likely the 10% estimate is too small, perhaps significantly so. We will assume that organizations that are not businesses might use 20% of the number of domain names utilized by businesses. Therefore we estimate the total number of domain names needed, both for businesses and other organizations, is of the order of 1.2 billion domain names.

What About Domain Phrase Needs?

Recently there has been interest in using domain phrases in marketing campaigns and redirections. For example Names.of.London and NamesThat.win both offer these services (and undoubtedly many others as well). It appears likely that creative uses of domain name phrases will increase significantly in the near future. Will this significantly increase the demand for domain names? The answer is probably not, since a single domain name is used with multiple first words, so the total number of new domain names is limited (essentially numerous sub-domains are used to support multiple phrases concurrently). While we see extensive growth in domain name phrase/expression use in the coming years, it will probably contribute only slightly to the total domain name market. Nevertheless, this is an example of the use of domain names in ways that were not foreseen even a few years ago.

What About Personal Domain Names?

In the near future will most people claim a personal domain name? If this were to happen, the potential market for domain names would be much larger, of the order of the global population of 7 billion. Clearly at the current time many people have no ability, or interest, in acquiring and using a personal domain name. Some promote the Montenegro country domain name (.me) as a personal website extension, and that TLD has grown in popularity driven largely by the personal domain market. There is a significant market for personal domain names, and it will probably continue to grow strongly in our opinion. However, we do not foresee a time when essentially everyone will have a personal domain name. We estimate, somewhat arbitrarily, that perhaps 1 person in 25 globally will want and have the resources to obtain and utilize effectively a personal domain name. This would suggest a need for roughly 280 million domain names for this purpose, or the total number of domain names needed globally for all purposes would be almost 1.5 billion by our analysis so far.

Future Uses for Domain Names

Will applications for domain names emerge that we don't even foresee now? For example, will one or more large retailers assign a domain name to each individual product, or will the ISBN system assign a domain name to each book product? Will advances in biotechnology, or other areas of science, lead to big data situations which are most efficiently handled by assigning a large number of new domain names? Will the structure of the Internet of Things emerge to assign individual domain names to each connected product? It really is not possible to meaningfully estimate the domain name market related to new technologies, but it probably makes sense to include some factor even though it is highly uncertain. We arbitrarily suggest that there might be a need for 300 million new domain names for use in ways not currently conceived and related to modern science, technology and innovation. Therefore we increase the total to 1.8 billion domain names.

How Many Domain Names are Currently Registered?

A recent report shows that the total number of domain names registered is of the order of 330 million domain names currently registered, with .com and .cn being the most widely held TLDs. Of that total, there are about 27 million new gTLD based domain names,

A good question is how many of these domain names are in active use, rather than simply being held for resale, and what fraction of these are potentially useful in meeting the market need for new domain names. One way to estimate this is to look at what fraction of domain names are parked. Among the
new gTLDs the parked fraction is typically a bit over 60%. The parked fraction is less for country code and the traditional gTLDs. Of the 330 million registered domain names perhaps 100 million are not actively used and it could be argued that many of these will not be attractive for future use and therefore do not contribute to the future domain name need. That is, perhaps 200 million useful domain names are currently registered.

How Many New Domain Names Are Needed?

Therefore this analysis suggests that a need for almost 1.6 billion additional domain names. This will be driven primarily by business uses (about 60%, even excluding new technology uses that are likely also business related), although non-business organizations, personal domain names, and future domain needs related to technological developments will all play significant roles. The market will decide how this need is met, through country code, traditional gTLDs, existing new gTLDs, closed TLDs, and not yet released newer gTLDs. We summarize in the next section the key results of this analysis.

Executive Summary

  • We estimate the number of businesses globally currently at about 100 million.
  • If on average each business utilizes 10 domain names, this suggests a business domain need for about 1 billion domain names.
  • Domain name needs of organizations that are not businesses might require about 200 million domain names, bringing the total to 1.2 billion.
  • The growing tendency for personal domain names may add another few hundred million to the potential market.
  • The use of domain name expressions or phrases will increase, but not add significantly to global domain name needs.
  • Future uses, not specifically predicted, related to advanced science, technology and commerce may add another few hundred million domain names.
  • This analysis suggests total domain name requirements of about 1.8 billion domain names when projected needs are realized.
  • There are currently about 330 million domain names registered, but probably only about 2/3 of those are potentially useful.
  • The analysis suggests the need for about 1.6 billion new domain name registrations.
  • If the figures in the analysis are valid, the number of domain names registered can increase from current levels by roughly a factor of 5 before saturation is reached.

Final Notes

This analysis is based on the concept of Fermi Problems, a technique popularized by physicist Enrico Fermi. He famously used the method to accurately estimate the number of piano tuners in the city of Chicago, and to estimate the energy of a nuclear blast from the displacement of a paper at considerable distance from the blast. The Drake equation, that estimates the number of intelligent extraterrestrial civilizations in our galaxy, is another well known application based on similar principles.

The technique cannot provide a firm number, but does provide a more reliable estimate than simply proposing a figure, and makes the assumptions of the components which go into an estimation. It would be possible to use
Monte Carlo methodology, based on uncertainty estimates for each component of the analysis, to express likely uncertainty ranges in the final result. Given the difficulty in assigning the component uncertainties, I decided that at the current time the precision of the analysis does not warrant an attempt to use Monte Carlo techniques to estimate overall uncertainties.

It is important to keep in mind that the paper estimates needs based on current numbers of businesses, but assuming that domain needs have been fully realized.

One of the biggest uncertainties in the analysis is the number of domain names per business. Should that number be at the lower end of our suggested range (about 2), this analysis would suggest that the total number of domain names needed is much lower, with business needs almost met by current domain name registrations.

This analysis is offered to the community without any implied or explicit warranty with respect to accuracy or applicability. We welcome comments and corrections to the analysis presented here.

The author, Dr. R. Hawkes, MSc, PhD, P.Phys., is a physicist and author who has employed data intensive analysis techniques in physical sciences. He operates several websites, including NamesThat.win, a site specializing in science, space and technology related domain name sales as well as use of domain expressions for marketing campaigns (Rent an Expression).

This paper is ©R. Hawkes, all rights reserved. If you would like to reprint the article please contact the author for permission and conditions. You may, however, quote small sections in news or review articles without specific permission, or reprint the executive summary as long as a link to the full white paper is also provided. You may also freely link to this site without prior permission.

Paper released July 31, 2017. This version released July 31, 2017.