Stringtel is a string discovery and risk-assessment tool that helps new gTLD applicants evaluate their chosen string -- or uncover other high-value opportunities.

Stringtel has two primary intended user bases:

• Organizations with a chosen gTLD, seeking risk analysis before applying.
• Organizations exploring potential gTLDs, looking for data-driven opportunities.

Stringtel implements the rules contained in the latest ICANN New gTLD Applicant Guidebook, using a database of over a million strings, to identify whether your search string risks:

• outright rejection (e.g., blocked geographic or organization names, technical issues),
• extra evaluation fees (e.g., high risk of name collisions),
• objections (e.g., from governments, blockchain naming system operators, or registries of visually similar TLDs),
• contention (e.g., with companies that have already announced applications, run the same string on popular blockchains, or applied unsuccessfully in previous rounds)

Even companies as well-resourced as Google had applied-for strings rejected in the 2012 round because they didn’t pay attention to the rules. Stringtel helps future applicants avoid these pitfalls.

Stringtel also does a broad string-similarity comparison between the user’s query and its database of protected and potentially risky strings.

Stringtel looked at over 180 million already-registered domain names in the most-popular legacy gTLDs. It knows how frequently every string is used an ending, immediately to the left of the dot.

If you know how popular a string already is as the ending of the second level, you can infer how popular it might be as a gTLD in its own right.

If there are 41,000 domains that already end in "bakery", for example, that may give you an idea of how popular a .bakery gTLD could be. Registrants of —bakery.com domains may opt to migrate to a shorter .bakery domain, and future registrants may choose to register a .bakery domain instead of a .com or local ccTLD.

As well as these "long-tail" niche opportunities, Stringtel can also help you identify common suffixes that could be used as "domain hacks".

The most-popular strings have Stringtel frequency values in six figures. Anything with a frequency below 1,000 is not counted.

In many cases, a string may only be popular because it is the substring of another string.

For example, 353,000 domains end in "port", but 141,000 of those end in "sport" and tens of thousands of others end in "support" or "report". That shows that .port may not be as attractive as a gTLD for sea ports as the top-line frequency value suggests.

Stringtel displays already-delegated gTLDs in red text and undelegated strings in black. In the search results for "port", we see that .transport is currently available and may be an attractive opportunity.

Any delegated or undelegated string may be clicked on to drill down into its own unique risks and opportunities.

Unchecking the “Show only dictionary words” box may reveal additional opportunities. For example, unchecking the box on the search results for "service" reveals "poolservice" and "cleaningservice" as commonly registered domain endings without currently matching gTLDs.

Other examples you can try out include "ology", "therapy", to discover which -ologies and -therapies are most popular. Searches for shorter strings such as "ing", "tor", or "use" could uncover even more opportunities.

Stringtel cannot tell you how many domains you can expect to sell in your chosen gTLD, but it can put your selected string in the context of other delegated gTLD strings and their frequency scores.

For example, .transport has a frequency score of 54,000, meaning 54,000 registered .com, .net and .org domains end in "transport". That's roughly the same score as the existing gTLDs .photos and .video, which have roughly roughly 23,000 domains under management (DUM) each and roughly 14,000 annual renewals.

Because a gTLD's volume could be affected by its age, pricing, and the marketing power of the registry behind it, hovering over the gTLD in the first column also reveals this additional context.

Your search term is compared against a database of roughly one million strings to check whether it would be banned by the rules of ICANN's Applicant Guidebook or attract additional scrutiny or contention.

For example, some strings are banned outright because they present technical risks, match geographic terms, or are plural/singular equivalents of existing gTLDs.

Other strings may present risks of objections or contention because they match, for example, already-announced applications, strings that have previously been applied for or attracted objections or strings that are commonly marketed on blockchain-based naming systems.

Stringtel uses several algorithms to determine the closest homographic/homoglyphic (looks like) and homophonic (sounds like) matches to other gTLDs or strings that could cause objections, contention, or other risks.

The algorithms are configured to provide a very broad set of matches, to help applicants narrow down the field of strings that may cause problems with their applications. In most cases, the results will be sufficiently dissimilar to not trigger an unfavorable similarity check. The results are no match for human common sense.