History is a great teacher, we are told. So, on the cusp of an explosion in new top-level domains, what can we learn from the two previous expansions of the Internet’s naming space? And what are the pitfalls to avoid?

Let’s just assume the fundamental and obvious lessons of realistic expectations, a solid business plan and prudent resource management, and instead focus on the little talked about but still critical lessons that will separate the winners and the losers in this race. But first—a caveat!

We are now in a different world when it comes to domain names. For one, regulatory changes will bring about an erosion of the traditional walls separating registries and registrars and open the way for new breed of registry/registrar entrants. Brands face complicated decisions on whether to block another entrant, apply for defensive purposes, or claim their own namespace as part of a larger strategic effort.

Instead of a handful of applications and one or two gTLDs being introduced over time, we could see hundreds of applications for this first round and a smaller subset of ASCII and IDN TLDs all vying for the same approximate launch window. In the meantime, new and alternative content navigation methods have captured the users’ imagination: content via apps (which users will gladly pay for), social media has exploded whether Facebook or Twitter. Search engines have morphed into intelligent and context sensitive advertising and productivity behemoths, and non-English speaking users are driving Internet growth, though not IDN usage.

So, what has remained constant? Dot-com is still king, or more specifically put, all of the TLDs launched since 2001 put together, have not managed to reach more than 20% of the total gTLD market, dominated of course by dot-com. So given all of these changes and a few constants, are there still reasonable lessons we can draw? Yes!

Lesson #1: Do not underestimate switching costs for the buyer. Recall the original aspirations and expectations of many TLDs who hoped to eat into the dotcom market share by convincing users to switch because “all the good names are available here!”

But which is more valuable: a loft in the Upper East Side, or a much larger and better-appointed loft in downtown Billings, Wyoming? As any good real estate agent will tell you, its location, location, location! It isn’t the loft per se, or in this case the name, it is the prominence of the neighborhood, the accessibility and the traffic that makes up the perceived value and hence underpins the desire to switch. So… dot-com remains king, while plenty of good names are still available in other TLDs.

The switching costs are often too high to do anything other than a defensive registration. Such registrations then suffer a much lower renewal rate as the registrants—who bought on a premise of fear—decide that the new TLD has not gained as much traction as they feared. So they drop the renewal. As a rule, switching costs for services that, have a limited number of suppliers, and are difficult to evaluate on a consistent basis, are high.

Lesson #2: Unless you have a simple explanation of a fundamental problem that your TLD is solving in a way that is faster, better, easier and cheaper, go back to the drawing board. TLDs have not gained momentum as way to certify a registrant or a website’s content, both of which we know are mutable from one instant to the other. It takes enormous resources to educate registrants of this intricate and complex value proposition—you need to create awareness, then turn that into a motivation to purchase AND create the content behind the site in such a way that it corresponds to the rules.

Lastly, there is an inherent tension between the registry’s economic incentive to keep as many registrations alive as possible, versus turning off those who do not comply with any stated criteria. Without a real value and a unique proposition for the TLD, the fall back position would be to sell registrations based on the need for brand protection alone. If this is the only strategy, then it is one bred by fear. And using fear to push sales has always been a shortsighted model, doomed to failure.

Lesson #3: The decisions you make in the very first stages of a gTLD launch are most predictive of its long-term success. A February 2010 study commissioned by ICANN [PDF] of gTLD launches since 2001, demonstrated that 40% of their lifetime registrations were achieved during the first year. There are critical decisions to be made in that first year: branding, pricing, marketing spend, policy and governance model, technical registry operations and setup, security etc. Any of these represents a variable, which if wrong, can poison the success of the TLD.

An added stressor is that once the TLD is launched there is no easy exit strategy. It is not like a product that can be pulled off the shelf (albeit with some brand damage and embarrassment). There is no easy exit strategy in the gTLD game. In other industries and with other products, if the initiative fails to meet expectations, you have the option to re-launch the product. Not here. The difficulty in reconciling the legacy market, policies and procedures, with new markets, policies and procedures is real and complex enough to increase the cost of a re-launch exponentially.

So hard decisions and trade-offs need to be made to make best use of limited staff, money and opportunity. For example: If you did not spend enough time, effort and money to create awareness of the TLD and stimulate a “pull response” from the channel in the months prior to launch, no amount of great relationships and channel incentives will ensure your longer term financial viability.

There are of course many other small but significant lessons. Curious as to what they may be? Then come to the .nxt conference in February to hear first hand the hard earned lessons from those of us who have been down this path before, and done so successfully.