The “hot topic” at the upcoming ICANN meeting in Singapore will, of course, be whether or not the new Top-Level Domains (TLDs) will actually launch or not. Sure, they’ll launch at some point, but ICANN has been pushing to make the big announcement at Singapore.

There has been a lot written over the last couple of years about new TLDs over the last couple of years. We are now coming into what might be called the “end game”.

However if the predictions on the number of new domain extensions are to be believed there are some issues that potential registry operators and ICANN need to be aware of. Unfortunately to date these issues seem to have been taking a “back seat”. I’ve tried to raise them in various fora, but I’m not overly convinced that prospective registries realise how much of an issue some of these things could be if they’re not dealt with sanely.

I’ve been considering writing about this for a few weeks and would like to emphasise one thing—in common with all my articles it reflects my personal opinion—no matter how flawed it might be.

From a registrar perspective I have very grave concerns that some of the applicants might be more than a little naive in their expectations.

Let me explain.

With or without vertical integration a domain extension’s success will depend on the marketplace. Put another way, if nobody registers domain names then your registry will fail.

It’s that simple.

Now if you’re lucky enough to have secured a highly desirable domain extension (i.e. objectively speaking there is an actual demand for it) then you might be able to afford to skip the rest of this article, but realistically speaking there probably aren’t that many domain extensions that will have guaranteed success.

Realistically speaking you will have to work with registrars, their resellers and the rest of the current market place. (Of course if the registry operator also operates a registrar with an existing customer base then a lot of these issues might not be that important) While you might think that your new extension and your business model is a “game changer” the hard reality is that it probably isn’t. Sorry.

So what things might you want to consider?

Let’s look at a few points to consider.

Selling domain names is like anything else. You need to remove as many obstacles from the sale as possible.

Registrars come in all shapes and sizes. Some of them are huge operations with hundreds if not thousands of staff. Others are small operations.

Who they sell to also varies considerably. While some focus on the “mass market”, others focus on the reseller channel, while others have chosen to work primarily with corporates. Others are only interested in specific geographic markets, while a lot of them probably don’t carry all of the existing top level domain extensions. Have a look at the list over on the ICANN site. As you scroll down that list you’ll see that while quite a few might have a reasonable number of extensions listed beside their names quite a few don’t. And I’m not even talking about how many of them have actually gone “live” with all the extensions—merely those who have actually signed the contract addendum with ICANN.

Put another way, how many registrars offer more than one or two of the more popular ccTLDs? Most stick to the most popular and the easiest to sell.

From a registrar perspective “going live” with a new domain extension doesn’t happen “overnight”.

The registrar has to sign the appropriate addendum for their contract with ICANN. Before they can do that ICANN staff has to check that the registrar is “in good standing” before they send out the necessary paperwork. The paperwork has to be printed out in duplicate, signed by an authorised officer of the company and sent back to ICANN’s HQ in Marina Del Rey. Even though ICANN has offices in other locations contract addenda are currently being processed in one location only. If there are 500 new domain extensions being added over the course of a few months then this process could get quite tedious and expensive, as you probably will use a courier to send back the docs if you’re concerned about them arriving or not. Once you’ve sent them to ICANN you then have to wait for them to process them. That could take several weeks.

It’s 2011. Why can’t ICANN process these contract amendments electronically? Why can’t they make them available via the registrar extranet OR provide some kind of “master” addendum allowing a registrar to be automatically opted in to ALL new extensions. Remember, the registrar won’t be obliged to offer them all, but at least they’ll have the option to do so.

Now you can get the paperwork from each registry operator. There’s going to be some form of contract and several other bits of paperwork to go with that. A lot of the paperwork will be almost identical regardless of which registry operator you’re dealing with, but you still have to complete the paperwork for each one separately and get it signed and sent back to them.

You then wait… and wait some more.

Again, this should be more automated. And how hard would it be for people to actually provide editable PDFs? Seriously.

Or, being logical about it, why can’t there be a uniform set of documents for all the technical, marketing and administrative details? Of course each contract will be different and you’d expect that, but forcing registrars to spend hours providing essentially the same information to hundreds of registries is a waste of time for all parties involved.

Eventually you get sent the technical details for accessing the registry’s backend in order to process registrations.

And this is where it could all fall apart really badly.

EPP is by its very nature “extendible”, so a LOT of registry operators add little “tweaks” to their implementation. While there might be some very good reasons for doing this every little “tweak” has to be handled by the registrars who want to offer the extension. That means development time. A “minor” change could lead to a major coding exercise for a registrar’s development team. Not only that, but if the registry has extra requirements at the time of registration then this can also impact how the extension is displayed in shopping carts etc.,

The registry’s registration policies should be as open and simple as possible. Of course, for some registry operators that are targeting a specific demographic there will be a requirement for stricter rules and policies, but unless a restriction is an absolute requirement then it should be removed. If you want to avoid issues then have a strong dispute resolution process.

Development time will lead to delays and headaches i.e. it’s an obstacle.

And what about the financial side of things?

Most registries work on pre-funded basis ie. registrars have cash on deposit with each registry operator. With each registration, transfer or renewal monies are debited from their account. If they don’t have sufficient funds with the registry they can’t process any registrations etc.

Pretty simple? Well it might be if you’re only dealing with a few registries, but if you have hundreds of them it can get quite messy.

Let’s say there are 500 new domain extensions. If the minimum funding level for each registry were set at 5000 (doesn’t matter which currency, but let’s use dollars), then that means that each registrar who wants to offer all 500 extensions will need to set aside 2.5 million dollars before they’ve sold a single domain!

That simply does not scale. Of course registrars are going to have to pick and choose which extensions they want to offer, but the funding level shouldn’t be the deal breaker. Of course you want to ensure that the registrars integrating with you are going to actively promote your extension, but if you consider the registrars’ overheads in terms of paperwork and technical implementation then it’s unlikely that many are going to integrate “for fun”. Not when you’re talking about that number of extensions.

There is no simple solution to this issue of course. Some suggestions to prospective registry operators:

• Accept credit card payments
• Offer “emergency” credits
• Set low minimum funding levels

At present, unfortunately, a lot of registry operators only offer wire transfers as a method of payment. That means that you can easily be waiting 2 or 3 working days for funds to appear in your account. While a registrar might need to be careful about the funding levels they maintain at each registry they interact with weird stuff happens. Based on my own experience I’ve seen a 300 or 400% increase in activity with a domain extension practically wipe out our balance in a matter of minutes or hours—and there’s no easy way to plan for that kind of crazy spike. Some registry operators offer credit card payments, which can help get you out of a “rut”, but a lot don’t. Some offer an “emergency credit” system, which can give you a bit of “padding” to make it through the 24—72 hours that it takes a wire transfer to hit their bank, but not all do.

In an ideal world there’d be a “clearing house” for the monies, but that’s not going to happen. However if a backend operator already has experience in handling the funding for multiple extensions, then maybe encouraging them to do that with the new extensions they’re handling might not hurt. For example Verisign, Afilias and COCCA already do this for several ccTLDs that they handle.

I already alluded to policy. I’m a very strong believer in the KISS philosophy. If you want people to be able to understand how they can register a domain name and use it then keep the policy as simple as possible. That doesn’t mean that you cannot have a strong policy. It just means that the policy should be worded clearly and simply. That means that everybody will understand what they’re dealing with and you’ll avoid headaches. If, for example, you’re going to be running multiple sunrises then make sure that the criteria for each one are as free from ambiguity as your lawyers will let you.

So moving past the “nitty gritty” above and assuming that we’re living in an ideal world where everyone is reasonable and wants to “play nice” together.

Up until now there have been one or two domain launches per year.

If you were to start looking at scaling that up to add say 500 new extensions in the first year then there’s a very obvious problem.

Marketing.

An average year has about 48 working weeks once you factor in holidays such as Christmas, Easter etc.,

If you want to launch 50 extensions in a year then that means that there will be more than one new extension per week. Multiply that to 500 extensions in a year and you’ve looking at more than 10 per week!

While of course not all new domain extensions are going to be focussed on the “mass market” it’s a pretty safe assumption that quite a few are.

How can you get your message across to the market effectively when you’re competing with so many other companies trying to do the same?

How do you get registrars (and other sales channels) to devote time, energy and resources to promoting your extension to their clients?

Most companies that send out email newsletters, for example, only send out one per month. How do you get them to promote the new extension for you?

Or are you going to do all the marketing and promotion yourself?

How about the people who are actually dealing with the end users? How are you going to make them aware of your new extension? Why should they even care?

There is no easy solution of course. If there was I’m sure someone would patent it and make lots of money in the process! (GoDaddy already has a patent for allocating “shelf space”)

There are some obvious things that potential registry operators should consider doing:

• Talk to existing registries
• Talk to existing registrars
• Be practical

The launch of new TLDs has the potential to change the entire internet ecosystem dramatically, but there is a real issue that many of the new TLD operators will fall at an early hurdle if they don’t give serious thought to how they’re going to actually sell their domain extension to the market.

Ken Hansen from Neustar wrote a post covering some of this which is also worth reading, as are Jothan Frakes’ slides from a CENTR meeting.