Alternative Perspectives on Registrar Market
Share - Revisited
The Fortune 1000, the Forbes International 500, and the Yahoo! Directory
Updated: October 2003
Benjamin Edelman - Berkman Center for Internet & Society - Harvard Law School
[ Overview - Data & Results - Summary Statistics - Discussion ]
Registrar market shares are measured in selected subsets of domain names, providing a basis for comparison with overall registrar market shares across the entire domain name market. Registrar market shares are found to vary dramatically across these subsets, with implications on the future customer retention rates of the corresponding registrars.
Since the 1999 introduction of competitive registration of domain names, analysts have examined the market share of the various registrars now able to offer registrations in the gTLDs of .COM, .NET, and .ORG. Initially Network Solutions (subsequently bought by VeriSign) served the entire market as the sole supplier of such domain names, but competitive registrars have subsequently come to have substantial market share. Understanding registrar market share is of interest in assessing the financial viability and proper valuation of each registrar, in selecting a registrar, and in understanding the registrar features and design seemingly of greatest interest to domain registrants.
Continued work by State of the Domain (among others) documents registrar market share as measured as a proportion of total COM, NET, and ORG domain names. In contrast, this document focuses on certain subsets of the domain registration market -- namely, the domains of the largest American firms (as listed in the Fortune 1000), the largest firms worldwide (as listed in the Forbes International 500), and the relatively well-known sites listed in leading web directory Yahoo!. These subsets of the gTLD domain name space are of interest for the distinct registration markets they represent. Most notably, this analysis focuses on domains actually in active use; the design of these subsets seeks to exclude domain name speculators and warehousers, whose millions of domain registrations might prevent ordinary domain registration analysis from properly reflecting the registrations of all domains actually in current active use and most likely to be renewed in the future.
Analysis shows substantial divergence among the groups studied. For example, while VeriSign's market share across all COM, NET, and ORG registrations has reportedly fallen to 26%, its market share among the primary domains of Fortune 1000 companies remains 74%, more than twice as large. Similarly, 62% of domains in a 2001 snapshot of Yahoo! are currently registered with VeriSign, and 38% of recent additions to Yahoo! use VeriSign. In short, VeriSign's market share is in every category of domains found to be larger than SOTD reports. Thus, excluding registrations by domain speculators and warehousers, it is likely that VeriSign has a larger market share -- of domains actually in use, for which renewal rates are likely to be relatively higher -- than SOTD's analysis has suggested. In contrast, certain other registrars tend to be underrepresented within these groups of registrants: iholdings.com, Schlund+Partner, and XIN Net are all underrepresented within the designated categories of domains, while experience shows them to be more popular among bulk registrants. The Results section of this document provides similar data for a variety of other registrars.
In this research, I document the registrars used by each of the Fortune 1000 companies, by each of the Forbes International 500 companies, and by listings in Yahoo!. I subsequently compare these registration patterns with the registrars used by the entire domain name registration market.
Data & Results
In recent testing and archiving, I have extracted a list of 997 distinct .COM, .NET, and .ORG domains used by 997 of the Fortune 1000 companies. (Two of the remaining companies lack official domain names in their Fortune listings, while one uses a .BIZ.) I have further prepared 277 COM, NET, and ORG domains reported by Forbes Magazine to be the primary domains of the respective firms in the Forbes International 500; many Forbes firms use domains outside COM, NET, and ORG, and since those firms' ccTLD registrations cannot be analyzed within the framework of gTLD registrar competition, they are excluded from analysis. Finally, I obtained two lists from Yahoo!: First, a random sample of 42,929 distinct registered .COM, .NET, and .ORG domains with at least one listing in Yahoo! as it stood in the spring of 2001. Second, 5239 distinct active .COM, .NET, and .ORG second-level domains that each host at least one web page added to Yahoo! between June 26, 2003 and July 25, 2003.
For each domain, I have used registry Whois services to obtain the registrar of record as of August 2003. A subset of the resulting raw data is available here:
Fortune 1000 companies with registrars - raw data
Forbes International 500 companies with registrars - raw data
I have tabulated each list of domains by registrar name to determine the number of domains registered to each registrar. The table below shows a portion of these results, while the entire listing of registration counts by category and registrar is available in a separate file. The "Overall" column of data comes from the Q1 2003 State of the Domain newsletter and reflects registrar market share across all .COM registrations.
|Registrar||Fortune 1000||Forbes Int'l 500||Yahoo!||Overall|
|Network Solutions, Inc.|| |
|Tucows, Inc.|| |
|Register.com, Inc.|| |
The listing of registrations by registrar is available in greater detail:
Registration counts by category and registrar
Registration counts by category and registrar, relative to overall registrar share
This data is graphically summarized in chart form:
Chart of registrations by registrar
Chart of registrations by registrar, relative to overall registrar share
Chart of changes over time, 2002 versus 2003
Data errata: registrars with multiple services
When a registrar offers multiple distinct registration services, whether due to acquisition of other registrars or due to internal development of multiple product lines, registry Whois data can be misleading as to registrants' specific registration choices.
This interpretative difficulty is most serious for registrations attributed in registry Whois data to VeriSign -- because VeriSign registrations could use VeriSign's Network Solutions registration service for small businesses and individuals, or could instead use Digital Brand Management Services (DBMS, a joining of registrars ID Names, 1GlobalPlace and NameEngine) for large companies. To take account of this problem, the author received from DBMS a listing of domains registered through DMBS, and using this data DMBS is listed separately from other VeriSign registrations in the Registration Counts tabulation above. In the future, when VeriSign completes its recently announced sale of its Network Solutions registration service, automated methods will again be able to determine registrar market shares, complete with basic information about type of registration service.
The author understands that Register.com also has a separate Corporate Services Division, but that Register.com's corporate registrations are typically automatically transferred to this division, eliminating the uncertainty of registration unit present under the VeriSign system.
Change Over Time
VeriSign remains the largest registrar in COM, NET, and ORG -- among Fortune 1000 firms, among Forbes International 500 companies, among Yahoo! listings, and among registrants generally. However, relative to similar analysis conducted in May 2002, VeriSign's market share has fallen in every category -- from 39% to 26% across all .COM domains, from 83% to 75% among Fortune 1000 companies, from 73% to 59% among Forbes Global 500 companies, from 74% to 62% among listings in Yahoo! as of 2001, and from 45% to 38% among new additions to Yahoo!.
Registrars with notable growth among Fortune firms include Register.com (9% to 12%) and Tucows (1% to 2%). Among Forbes firms, Register.com also grew (8% to 12%) as did Tucows (2% to 3%) and Melbourne IT (3% to 5%) and CORE (3% to 4%).
Among Yahoo! registrations, growth took place at Go Daddy (old Yahoo! list: 1% to 3%; new: 3% to 4%) and eNom (2% to 4%, 2% to 5%).
Throughout all subsets of domain registrations, as well as all .COMs taken as a whole, the largest registrars are jointly falling in market share, while smaller registrars are growing -- on balance an increase in competition. For example, in 2002, the top 10 .COM registrars operated 80% of .COM domains, while in 2003 this figure has dropped to 76%. Among Fortune firms, the drop is from 85% to 91%; among Forbes firms, from 87% to 80%; among Yahoo! listings as of 2001, from 93% to 88%; among new additions to Yahoo!, from 88% to 82%.
Fortune 1000 Registrations
Fortune 1000 domain registrations continue to show an apparent preference for VeriSign. In particular, VeriSign serves nearly three times as many Fortune 1000 registrations (75%) as its share among registrations generally (26%). An additional 12% of Fortune 1000 domains are registered to Register.com, an amount slightly larger than Register.com's overall market share of 9%. No other large registrar (with overall market share above 1%) has Fortune 1000 market share of more than half of that registrar's overall market share.
Some small registrars serving the Fortune 1000 serve significantly more Fortune customers than their overall registration counts would have predicted. For example, Alldomains.com serves only 0.36% of all .COM registrations, but it serves fully 3% of Fortune 1000 firms. Though Fortune market shares are smaller for other firms, both MarkMonitor and NameEngine (now part of VeriSign) also serve disproportionately more Fortune firms than their ordinary registration shares would have predicted. These results may reflect that certain registrars offer premium services (such as monitoring) that are of particular interest to large corporate customers.
Among large registrars, several show apparent disfavor among Fortune registrants. For example, Tucows has 10% of the overall .COM market, but only 2% of Fortune registrations, and similar patterns hold for Go Daddy, eNom, Melbourne IT, and Bulk Register. These results suggest that, through the services these registrars offer, they have made themselves significantly more appealing to ordinary and high-volume registrants than to Fortune 1000 registrants.
Taken as a whole, the Fortune 1000 registration market is becoming increasingly consolidated with two large registrars, VeriSign and Register.com -- the former slightly falling in Fortune 1000 market share between 2002 and 2003, the latter growing. Most of Register.com's growth came at the space of smaller registrars, which in 2002 jointly held more than 9% of Fortune 1000 registrations but in 2003 hold less than 5%.
Registrar choices remain roughly equal among Fortune 1000 companies with increasing ranks versus those companies whose position on Fortune listings has fallen in the past year. However, Corporate Domains shows increased popularity among companies with rising ranks (1.3% share) versus among those with falling ranks (0.5%). In contrast, Go Daddy has noticeably greater share among companies with falling ranks (1%) than among those with rising ranks (0.4%).
Registration of Fortune 1000 domains shows a split between domains under management by a registrars specializing in corporate domains versus ordinary registrars serving customers of all types. Registrars specializing in corporate domains include VeriSign Digital Brand Management Services, Register.com Corporate Services Division, AllDomains, and MarkMonitor, which respectively manage 17.54%, 11.72%, 3.01%, and 2.51% of Fortune 1000 companies' primary domains. The remaining 65.22% of Fortune domains are managed by ordinary registrars. (Note that MarkMonitor and AllDomains on October 20, 2003 announced their plans to merge.) The chart at right shows shares of Fortune 1000 domains registered by registrars specializing in corporate registrations.
Forbes International 500 Registrations
Among Forbes registrations, VeriSign's market share was 59%, somewhat lower than its share among Fortune firms. Register.com's share remained similar, at 11%, but Tucows, CORE, Melbourne IT, and Transpac showed notably greater share among the more international Forbes firms than among the American firms listed by Fortune. This may reflect a preference among non-American firms to use non-American registrars, i.e. registrars in their geographic region.
Registrations of Domains Listed in Yahoo!
A snapshot of Yahoo! from 2001 shows an allocation of registrations among registrars that again strongly favors VeriSign. VeriSign has a 62% share of such domain registrations, followed by Tucows with 6% and Register.com with 4%, and Go Daddy, eNom, Bulk Register, Melbourne IT, Dotster, Directnic, and Domain Discover, each with more than 1%. Relative to data collected in 2002, this reflects a drop for VeriSign (from 74%), large gains for eNom and Go Daddy, and modest growth for most other large registrars.
New additions to Yahoo! also reflect a shift away from VeriSign. Only 38% of domains added to Yahoo! in June-July 2003 use VeriSign as their registrar, a drop from last year's 45%. Among large registrars, Go Daddy and Melbourne IT are slightly overrepresented in new additions to Yahoo!, while Schlund + Partner, CORE, and dotregistrar are notably underrepresented, each registering less than a third as many new additions to Yahoo! as their overall COM registration rates predict.
Residual Domain Registrations
While the categories described above do not exhaust all substantive uses of domains, they are generally representative of domains that are in active use providing substantive and unique web content. Of the categories proposed, those with at least one listing in Yahoo! are likely most representative of domains in active use. Accordingly, registrars that are underrepresented in Yahoo! -- with fewer domains listed in Yahoo! than the registrars' overall registration rates would predict -- are likely to tend to register domains not subsequently put to active use. Such registrations include domains held for defensive registrations, domains used for warehousing, domains that capture residual traffic from others' registrations that have lapsed, and domains that provide sponsor links.
Registrars with notably low presence in Yahoo! registrations -- and therefore with notably high registrations of domains not in active use -- include iholdings.com, Schlund + Partner, XIN Net, Global Media Online, Yesnic, Nordnet, and Active ISP.
Discussion, Policy Implications, and Future Work
Fortune 1000 Registrations
Although the primary domains of Fortune 1000 companies represent a very small sample of all gTLD domain registrations, these and other major corporate registrations may represent a disproportionate share of registrar profits. Corporate customers may tend to select premium registration services associated with higher fees than those ordinarily paid by individuals, small businesses, and speculators. Further research might attempt to better quantify a wider notion of a registrar's "premium customers" and subsequently examine which registrars have the most such customers. Of course, to the extent that even Fortune 1000 companies purchase simple registration services from their registrars (perhaps because their value-added work takes place internally or at a law firm or other provider), these companies may not in fact be "premium" customers.
The Fortune and Forbes analysis considers only the primary domains used by the respective firms, as reported by the respective publications. Fortune and Forbes firms are likely to hold multiple domain registrations, and it would be desirable to examine trends in these many registrations. For example, examining this data would determine whether VeriSign continues to maintain notably high market share among Fortune firms' additional domain names, or whether its large market share is limited to Fortune firms' primary domain names.
The clear divergence in registrars' focus customers confirms an anticipated benefit of registrar competition. Some registrars focus on premium corporate customers while others rely on the mass market -- precisely as contemplated during design of the competitive registrar system.
Forbes International 500 Registrations
The apparent preference among non-US firms for non-US registrars confirms an anticipated benefit of registrar competition -- that registrants can do business with a firm in their geographic area.
Registrations of Domains Listed in Yahoo!
Registrar operations benefit from high domain name renewal rates: With high rates of renewal, a registrar earns substantial predictable recurring revenue, while lower renewal rates require that a registrar find new customers to meet revenue targets. It is likely that the domains listed in Yahoo! will have higher rates of renewal than other domains due to the relative popularity of these domains, their recognition by Yahoo!'s editors, and their prominent promotions in the Yahoo! directory. Accordingly, even as domains listed in Yahoo! may offer only average revenue per registration, a registrar with many Yahoo! domains may see higher renewal rates in the future and therefore higher revenue in the long run.
Further Implications on Evaluation of Registrar Competition
This research quantifies outcomes that suggest significant benefits to registrar competition. Identifiable groups of customers are found to choose particular registrars that seem to address their particular needs; for example, some large firms choose registrars with special monitoring services, and non-US firms disproportionately tend to choose registrars located outside the United States. In addition to serving the market through lower prices and higher quality, these new choices likely lead to better matching of suppliers with customer needs -- further increasing consumer surplus. In related research, a Survey of Domain Registration Services, the author is attempting to investigate, organize, and quantify the differences in services offered by registrars -- and to determine whether such service differences are in fact significant predictors of registrar choice.
Characteristics of the addition of new sites to Yahoo! suggest that registration of new domains is well subject to competition. However, there remains a substantial divergence between the registrars of new additions to Yahoo! and the current registrars of sites long listed in Yahoo!. This divergence might be taken to imply a low rate of switching from the registrars previously found to be optimal (perhaps in a time of lesser or no registrar choice) to the registrars currently thought to be optimal for new registrations. Registrants' decision to stay with, in general, VeriSign may reflect registrants' rational evaluation of the benefits and savings associated with transferring a domain to a new registrar versus the costs and risks of doing so. To the extent that registrant decision-makers have imperfect knowledge of the facts at hand -- for example, to the extent that they perceive large risks where risks are small -- corrective communications are likely to be necessary to assure efficient decision-making. To the extent that transfer procedures are in various ways difficult, confusing, or poorly-documented, a reduction of such costs and confusion (accompanied by a widespread publication of such improvement) would likely encourage additional transfers and increase the rate of equalization between the "snapshot" and "new addition" registrar market shares.
This analysis considers only market share within designated gTLDs, an important indicator of registrar status but not the only such indicator. For a full picture of registrar health and long-term prospects, it would be desirable to consider, and to quantify, other markets and other services of interest. For example, most large registrars now offer at least some ccTLD registration services, and these services may be particularly profitable since they offer additional value-added when ccTLD registration is more difficult than in gTLDs. Certain registrars have also expanded to offer additional services of particular value to large registrants -- including workflow systems to streamline multistep registration and approval processes, as well as legal analyses to assess trademark conflicts and evaluate approaches to dispute resolution. The author's Survey of Domain Registration Services attempts to systematically track some of these new services.
Last Updated: November 21, 2003 - Notify me of major updates and related work.
This work builds on and supersedes my earlier "Alternative Perspectives on Registrar Market Share" (2002) and "Fortune 1000 Domain Registrations: An Alternative Perspective on Registrar Market Share" (2002).
This research is supported in part by VeriSign, Inc.
This page is hosted on a server operated by the Berkman Center for Internet & Society at Harvard Law School, using space made available to me in my capacity as a Berkman Center affiliate for academic and other scholarly work. The work is my own, and the Berkman Center does not express a position on its contents.