Philip Corwin image

New gTLDs: Competition or Concentration? Innovation or Domination?

Philip Corwin imageThis guest post was written by Phil Corwin and originally appeared on DomainNameNews.

Expect the unexpected. Because it will happen. And it has just happened in the application phase of ICANN’s new gTLD program, with potentially profound consequences for the future of e-commerce.

During the three year period between the June 2008 ICANN Board approval of the new gTLD program and its June 2011 vote to proceed to the application stage, and even beyond then in the context of continuing GAC-Board discussions, only one competition issue ever became the subject of heated and protracted debate. And that was whether ICANN’s requirement for registry-registrar separation should be relaxed in concert with the new gTLD program, a question that ICANN eventually answered in the affirmative notwithstanding resistance from some members of the GAC.

But the consequences of that decision will hardly be a ripple across the Internet. While the consequences of what has happened could be a landscape-sweeping tsunami. And that is because the dominant search engine, Internet data aggregator, and Internet advertising provider, along with the leading online retailer, have emerged as the number two and three applicants for new gTLDs. Not just for their trademarks/brands, but for dozens of key generic terms. And, in a process that culminates in auction whenever any other non-community applicant for the same ‘string’ elects to stay in the fight, they have the bulging wallets to shove others aside and acquire their desired gTLDs. If they secure their bids the nature of Internet search and commerce could be permanently transformed in a manner far afield from the ‘competition and innovation’ mantra used to justify this massive roll of virtual dice.

The reaction has already started. In an article published in the June 14thWashington Post just one day after ICANN’s “Big Reveal” event in London, titled “Google, Amazon lead rush to secure dot-dominance”, the perceived stakes were described:

Amazon and Google are staking claims to large swaths of the Internet under a new system for labeling Web domains, bolstering their ability to control traffic as the Web expands beyond the realms of “.com,” “.gov” and “.org.”

The bids by those companies to acquire new domain names such as “.book,” “.shop” and “.movie” renewed fears among competitors that a powerful few will dominate the Internet marketplace of the future…

If Internet users embrace the new domains, the companies that control them could bear considerable influence on Web traffic.

Amazon has applied to control the “.book” and “.movie” names, for example, meaning that anyone else selling those items would have to get the company’s permission to be listed within that domain.

The National Retail Federation had urged that oversight of such generic domain names be given to impartial entities rather than individual companies.

“The results for now are as potentially unfair to businesses and consumers as we feared they might be,” said Mallory Duncan, general counsel for the trade group…

Duncan said consumers may not realize that the new domains are under private control and that the open competition that prevails within the “.com” realm may not exist within, say, “.grocery.”

“Consumers going to that domain may not realize that all of their shopping is being done with one company instead of a competitive market,” Duncan said.

Google was among the most prolific applicants, seeking to register 101 names at an application cost of $18.7 million. Never lacking in its quest for virtual completeness, the company is seeking to control “.mom,” “.dad” and “.kid.”

Amazon applied for 76 new names, including “.amazon” and “.zappos.”…

Others, however, are bracing for the giants of the Internet to seize even more power over its commerce.

“It would be wrong on so many levels for Amazon to acquire either the ‘.book’ or ‘.author’ top-level domains,” said Paul Aiken of the Author’s Guild. “Their ambitions to extend their monopoly in bookselling have long been abundantly clear, and with their cash, their technical knowledge, this could be yet another way in which they’ve extended their control over the book market. This really makes no sense.” (Emphasis added)

It does not matter whether one agrees with the concerns expressed by the National Retail Federation and the Authors Guild. What matters is that they and many other interests who have expressed strong misgivings about the new gTLD program have been handed a new and volatile issue by the unexpected length and scope of the Google and Amazon applications.

The potential effects on competition are major yet quite different for these two Internet giants.

Google is the massively dominant Internet search and advertising firm. One must assume that its applications for dozens of generic gTLDs beyond its brands (such as .gmail and .youtube) are based on proprietary analysis of the trillions of search terms typed into Google since its inception, and a conclusion that these gTLDs are most closely correlated with income-generating consumer search patterns.

Understanding the implications of Google’s bid requires a basic familiarity with the two species of Internet search. Google already dominates mediated search, in which its ever-morphing “secret sauce” algorithm takes a user’s search term and then presents a long list of websites deemed to be relevant. So important are these results that an entire SEO-optimization sector exists to assist online businesses in maximizing their odds of being displayed on the first few pages of a Google search.

The other species is direct search, the approach long used by domainers who seek to acquire and monetize noninfringing generic word domains capable of attracting type-in traffic at .com and other leading gTLDs, as well as well-trafficked ccTLDs like .uk and .de. Some members of ICANN’s Board, while disabused of the notion that all domainers are cybersquatters, may still view them as somewhat unsavory speculators rather than savvy investors. Yet, ironically, the entire new gTLD program is based on the direct search model pioneered, refined and monetized by domainers – elevated from the second level of the DNS to the first – and entrepreneurial new gTLDs applicants are investing/speculating in the belief that controlling key generics at the top level is a winning business plan. (Likewise, the “digital archery” methodology selected by ICANN to determine batching queue status for new gTLD applications bears close resemblance to the “drop catch’ techniques used by domainers to acquire expiring domains re-entering the secondary market.)

So the 101 gTLDs applied for by Google carry the possibility that it can expand its mediated search dominance into key direct search gTLDs, thereby further consolidating its control in connecting consumers with e-commerce while adding to its invaluable proprietary database as it peruses new gTLD registry data that no one else can access. It is questionable how much new competition will result from the integration of more virtual real estate into the Googleplex-dominated cyber-empire and the accompanying expansion of its Internet hegemony. There may also be domain marketplace spillover effects in terms of Google having a registration advantage over other new gTLDs that might be used to similar ends by the same registrants, if there is any perception that domains at new Google-controlled gTLDs will somehow gain favor in Google search engine results over the same domain registered at a competing non-Google gTLD. It is also unclear whether Google will make the most desirable second level domains at its new gTLDs available to the general public – other new investor-backed gTLDs will almost surely elect to sell the most coveted second level domains to the highest bidders during their ‘land rush’ period, but Google’s wallet is so fat that it can retain for its own purposes all second level domains predicted by its crunching of search data to be most trafficked.

The issues are somewhat different with Amazon’s bid for 76 gTLDs, including .book, .cloud, .game, .movie, .play and .shop. Again, one presumes that Amazon based its selections in large part on its proprietary database of searches done by shoppers looking for goods on Amazon. Competitor e-retailers would have understandable trepidation about registering their own websites on real estate controlled by the leviathan of Internet retailing. But, somewhat incredibly, Amazon’s applications all carry the notation that “Amazon and its subsidiaries will be the only eligible registrants”. While this would be understandable for gTLDs like .amazon, .kindle and .fire, it becomes more problematic for dozens of generic word applications. While Amazon might open them to public domain registrations down the road, for now at least their object seems to be the assurance that competitors cannot acquire them and use them as a platform for competing retailer services within a vertical generic silo.

If we liken the new gTLD program to a land rush, this would be equivalent to the U.S. opening the Oklahoma territory to farming and ranching in the 19th century and incumbent agricultural interests acquiring the most valuable acreage to make certain it continues to lay fallow, thereby preventing any lowering of commodity prices. This is indeed ‘squatting’ at the highest level of the DNS, and how Amazon’s ‘No Trespassing’ plans translate into any new competition or innovation is near-impossible to discern. (Note: Google has similar plans for some of its generic bids – such as .cloud (again!) and .blog – with the seeming aim of acquiring the new gTLD primarily to prevent anyone else from developing it through domain registrations open to the general public.)

Adding to the policy mix is the fact that ICANN has structured the new gTLD program without qualitative analysis of applicant intent outside of the narrowly focused objection and public comment process, and with no requirement that any new gTLD be developed to even a minimal extent. All applications that survive that gauntlet eventually go to auction if competing bids do not drop out, agree to be bought out, or reject an invitation to join forces. For the 21 new gTLDs for which Google and Amazon submitted competing bids (based on exact match or confusing similarity) — including .app, .cloud, .free, .search, and .store — the competition will be a bit like Godzilla vs. Rodan; as of March 31, 2012 Google’s cash on hand was a towering $49.3 billion, while Amazon’s was a relatively paltry but nonetheless impressive $2.3 billion (and this does not take into account that competitors can be persuaded to drop out or merge efforts by grants of stock in the parent bidder or the new gTLD registry). For all other applicants competing with Google or Amazon the results of that contest will be similar to the faceoff between Bambi and Godzilla – splat! road kill!

There are of course other applicant groups who have put in bids for scores of new gTLDs. However, none of them represent the dominant search/advertising and retailing companies on today’s Internet, so they can make a better argument that their acquisition of multiple new gTLDs may actually introduce some effective new competition and innovation against the dominant players. And it is doubtful that any of them plan to let their new gTLDs lay fallow – their aim is likely to maximize domain sales to and usage by registrants seeking quality domains in relevant generic silos.

The competition issues presented by these large and unanticipated gTLD bids from Google and Amazon may not fit neatly within traditional antitrust analysis but are important nonetheless, and compliance with ICANN’s rulebook is no assurance of adherence to relevant national laws. The U.S. Supreme Court has defined monopoly power as “the power to control prices or exclude competition” in a particular market — which raises the issue of whether the relevant market is the universe of all gTLDs or a single generic vertical gTLD silo. The Sherman Antitrust Act makes it a felony for any person to “monopolize . . . any part of the trade or commerce among the several states, or with foreign nations”. The elements to be proven under the Sherman Act are that the defendant possesses monopoly power in a properly-defined market and obtained or maintained that power through conduct deemed unlawfully exclusionary. Again, how these principles might be applied to new gTLDs is unknown territory – just the kind of cutting edge issue that aggressive prosecutors like to explore and lay claim to. It is clearly arguable that if Google and Amazon are each allowed to own all of these domains and to use them to promote their own products and services, or to prevent others from making them available for public domain registrations, that it will be exclusionary to all of their competitors. The remaining question would be whether such conduct is unlawfully exclusionary, and that is a question that could take years of litigation to answer.

We are less familiar with EU competition law and policy, but our impression is that it is broader and more flexible than the U.S. approach and thus even more likely to provide a context for raising questions about these gTLD bids.

As for ICANN’s internal review process, potential objections to the Google and Amazon bids don’t appear to fit neatly within any of the four objection classes that ICANN has established for the program and that will be open for the next seven months, so we will most likely see initial competition-based objections surface during the 60-day post-application public comment period.

Of course, the real pressure may be brought upon ICANN as groups with concerns about the Google and Amazon bids (including, perhaps, other new gTLD applicants) meet with legislators and regulators in Washington, Brussels, and other national capitals to urge investigations, hearings, and intervention. Google has already been the recipient of much antitrust scrutiny, especially in the EU, but also from the U.S. Federal Trade Commission which has also been actively monitoring ICANN generally and new gTLDs specifically.

Such legislative and regulatory activities could in turn trigger objections from the GAC, which has instructed ICANN that it reserves the right to oppose applications that are not in the ‘global public interest’ – and which is not confined to basing objections on pure statutory law. At a time when ICANN’s preservation against potential takeover of its key functions by the ITU or other entities is a very live issue, and when its re-award of the IANA function contract is still in limbo, it could ill afford to ignore such protests from governments otherwise supportive of its multi-stakeholder governance model. Especially when those governments have other more pressing concerns, such as staving off currency collapse and economic contraction, and will have little patience for controversies that might well have been avoided with better planning.

Should such protests arise against the Google and Amazon bids, it is unclear how that would impact ICANN evaluation of competing bids for the same gTLDs. If the two companies dig in and insist that their generic gTLD bids will not have an anticompetitive effect, and refuse to withdraw them, the processing of all competing applications for identical and similar strings could be put on hold until any relevant regulatory actions or litigation are resolved. Which could be months or years.

So, to circle back where this article began, these unexpected bids from Google and Amazon could have unexpected repercussions for the entire new gTLD program and for the future development of the Internet and e-commerce. Stay closely tuned as more unanticipated developments are almost sure to ensue.

As mentioned above, this guest post was written by Phil Corwin and originally appeared on DomainNameNews. Mr. Corwin is Founding Principal of the Virtualaw LLC consultancy and serves as Of Counsel to Greenberg & Lieberman and as for the Internet Commerce Association (ICA), all located in Washington, DC. This post is his personal opinion.

The article has been reproduced with the permission of Phil Corwin and DomainNameNews.