I think it's difficult.
Hundreds of corporations have applied for their own strings so they need to start using them.
Influential companies like Google can put gTLDs in the public eye, but so far they have been timid between domains.google and abc.xyz, that nobody visits.
So, if big companies start using (and
advertising) their new strings prominently, the public will take notice. Unfortunately, that doesn't mean people will want to buy a .xyz or a .gdn just because Google are using .google. People have no idea of the choices available to them, the TLDs that exist and those that don't (yet).
If you try to register a domain at Godaddy you get plenty of suggestions for new extensions. Unfortunately, many are not relevant, because the TLD doesn't match the keyword. And you only get a selection, they cannot display all the possible extensions to your face. Another problem is that many TLDs are more expensive, sometimes
very expensive if you try to register generic keywords. It's a real put-off to end users.
Maybe a good example of promotion would be the Godaddy Super Bowl ads for .co. But it's been a one-off experience. Perhaps the effort was not worthwhile ?
As a domain seller I like to say that the domain must speak for itself, then it will sell. The same logic applies to TLDs. They must at a minimum be
relevant to the target group. Too many strings are pointless, so there is nothing you can do I'm afraid.
Perhaps TLDs could differentiate themselves with certain services/features. For example, in the very early days when .pro was highly restricted, it was bundled with SSL certificates.
Maybe some community TLDs could appeal to specific populations, like .cat.
The sad truth is that while Icann are pocketing huge sums of money, the registries have invested a lot and they are having a difficult time. I don't think many are profitable at this point.
Marketing is expensive and few TLDs have the potential to muster enough registrations to make it worthwhile. So many TLDs end up in auto-pilot mode...
Some geoTLDs are promoted a bit, but with taxpayer money.
In many instances new extensions will not be an alternative to .com but additions: many will be used as redirects or for satellite sites, or time-sensitive campaigns thus they will more often be used as disposable items.
So far I think it's the registrars that do the marketing. They present options to their customers but also send out newsletters. So anybody who already has a registrar account must have heard about new extensions. They were even featured in the mainstream media. Registrars also advertise in the magazines for hosting, but for domains too, promising an abundance of new choices of domains, that will often not materialize
Registries have released press releases on the occasion too.
