New pioneering ideas are not always welcomed or requested by consumers. I outline three such examples:

1. The car. “If I had asked my customers what they wanted, they would have said a faster horse,” Henry Ford once said, or so goes the story. The quote has never been verified, but what it says is true. In 1903 most consumers gave little sign of wanting to buy cars. The Ford Model A began rolling out and all that changed.

2. ATMs. In the late 1970s, many people were still suspicious of impersonal machines. Customers were used to having their money handed to them by a teller at the bank. With an ATM they had to stand at a machine on the street and punch in a code. Citibank took the leap despite consumer sentiment, not because of it. But in 1977 the bank invested about $160 million to fill New York City with ATMs. In a few years, what people thought of as banking had changed.

3. Amazon. Jeff Bezos launched the company in 1994. As of 1995, according to a survey by the Pew Research Center, 97 percent of Americans had no online access and 42 percent had “never heard of” the Internet. Soon enough they did.

Henry Ford didn’t wait until people wanted a mass-market car. Citibank didn’t wait until they wanted ATMs, and Jeff Bezos didn’t wait until they decided to shop on their computers. But the three innovators did more than follow a vision. Before they made their respective leaps, they pondered what people did want. They asked themselves some hard questions.

Determining what consumers need

What are the painful consumer problems that the above innovators addressed? The real problems that the introduction of cars would have solved include flies, disease, smell, dried manure dust, cruelty to horses, and horse-related traffic deaths [today, of course we have traffic deaths caused by smartphones :( ]. The ATM innovation made banking available 24/7 and reduced customers’ time waiting in line. With technological advancements, mobile banking has further increased global convenience and accessibility. Amazon’s online bookstore introduced a personalized book-recommendation system, made available rare books that bookstores don’t typically carry (long-tail items), made online credit card payments acceptable, and provided delivery that was especially valuable for remote rural area residents.

It looks like ICANN and the registries skipped that step with the new gTLDs. Here’s the first question that should have been asked: “Will businesses need gTLDs to alleviate any pain caused by unmet needs, or do they just want the things?” ICANN didn’t bother. The thought of a cash cow drove out analysis.

But a successful innovator must figure out what customers will buy. That means thinking about what makes their lives less painful. Looking at the prospect of new gTLDs, ICANN should have asked whether businesses needed them for branding, labeling, or both. From that, a few other questions follow. Will the domains make life easier for Internet users? Is there a need for .brand? Why did earlier introductions of gTLDs fail? Assuming that new gTLDs are needed, will they create social value? ICANN didn’t ask these questions or any others.

The registries failed to ask some questions too. Of course, the new gTLDs weren’t their idea. ICANN had made up its mind and commissioned experts to support its view, sometimes with rather dubious studies. But the registries should have done a better job of assessing which gTLDs they decided to buy. They should have investigated which, if any, new gTLDs were needed for branding and which, if any, for labeling. They should have asked whether there were better substitutes that competitors might apply for. Basically, they should have tried to minimize the chance of ending up with a second-best gTLD that didn’t meet customer needs. They should have asked whether any given gTLD would reduce the pain for businesses and Internet users.

Here’s the second fundamental question that didn’t get asked: do the risks outweigh the benefits? That question should cut especially deep with ICANN, since the group is a monopoly and therefore has a special responsibility toward Internet users. Some follow-up questions: Would introducing more than a thousand new gTLDs cause technical instability in the Internet? With .com no longer a default, would Internet users be confused as to which was a business’ branding domain and which was its labeling domain? Would the domains be adopted at all? And what would happen to owners of complementary assets? After all, a registry failure impacts more than the one company going belly-up.

A question the registries skipped: what about the risk of user confusion? The spate of new domain labels would include many that were close substitutes for each other. Labeling keywords could have singular or plural versions. And, by the very nature of the change, consumers would have to start guessing whether a company used .com or something unfamiliar and novel. Would the benefits of the new names outweigh the risk of users throwing up their hands?

Where do you find the answers?

The best and perhaps only way taking a calculated risk is to talk to customers. ICANN didn’t; fixated on its chance at a long-running income stream, appears not to have thought too hard about the program’s viability. As for registries, it is not clear if they asked their customers, not when registration numbers are dismal and renewal rates for some gTLDs fall below 70%. Incumbent registries had an advantage over newcomers when it came to getting in touch with customers. They didn’t use it because they didn’t know there were questions to be asked.

Highly skilled venture capitalists (VCs) have a success rate of just 10% or so for the projects they finance. With that in mind, the temptation for new gTLD registries is to apply for plenty and hope that a few will succeed. But success rates of VC projects are only marginally correlated (i.e., the success rates of projects are practically independent of each other), compared with those for new gTLDs. Thus, the entire program can fail even if a few gTLDs look like they’re succeeding for a while. This is another reason to limit the number of new gTLDs, in addition to experts’ public warnings and suggestions, including my 2010 post.

In conclusion, there are few lessons to learn from this round, save some tweaking of auction rules and ICANN eligibility requirements. But round two will be full of surprises. As I have noted in a previous essay, successfully picking industry labeling and branding competitors to .com gTLDs will be a lucrative investment, albeit at a higher risk than an average new gTLD. In addition, there’s a new frontier in gTLDs that reflect house services (such as electricians, plumbers, dog walkers, and other manual labor). But remember, in contrast to popular belief, studies show that relying on success factors and analyses of what these factors have in common is an unreliable predictor of success! With such a high risk of failure to the Internet navigation structure, we could have been better off with some oversight from governmental and/or international organizations.