Thinks AI will impact domain names — How Verisign
Verisign’s CEO talked regarding the impacts on the prior week’s earnings conference call.
On DNW Podcast #545, I discussed how AI is impacting domain names with Carlos Armada of name.com.
What does Verisign, the .com and .net registry, think the impact will be?
Officially, Verisign’s SEC filings mention AI as a risk factor:
The number of domain name registrations under our management may be negatively impacted by certain factors, counting overall economic conditions, competition from country code top-level domains (“ccTLDs”), other gTLDs, services that offer alternatives for an online presence, such as social media and artificial intelligence (emphasis added), and ongoing changes in the internet practices and behaviors of consumers and businesses.
But aside from SEC disclosures, what does the firm think AI might mean for domain demand?
Verisign also sees upside. On the prior week’s investor conference call, Verisign CEO James Bidzos stated:
And with respect to AI, this is really interesting…as I’m sure you are, I get a lot of questions, a lot of interest. It’s really the focus for everybody. We don’t see it as a negative. I think there are at least 2 areas where AI can be positive for domain names. First, AI models create their data from websites, and we think the AI optimization will become more notable for creating websites with more content. I’d like to think that there is no static data in the world today even if an AI had an LLM that consisted of all, everything in the Encyclopedia Britannica, even the War of 1812 is getting a fresh look, a fresh analysis and fresh information. So there’s nothing static. So information needs to be obtained and the Internet needs to be navigated and that’s where our infrastructure comes into play.
AI tools can also be powerful for domain name suggestions, really powerful. And we’ve learned over the years that multiple keyword coms are good domain names to have. And we’re using this building it into our name suggestion tools.
There are a couple of things to unpack here. The first part of his statement seems to allude to how sizable language models need to suck in fresh information to provide accurate information. This fresh information will come primarily from websites.
While true, I wonder if the supply side of creating and publishing online will decline as the economic incentives change. Traditional search is driving fewer visitors to websites, and AI searches are not filling in that gap.
The prior week, I received a notice that one of my hosting plans was regarding to renew for the year. It was for a site I created a few years back that is entirely reliant on search traffic. It’s a long-tail site that can’t rely on repeat visitors, and the information it offers, although originally researched, is precisely the type that AI tools excel at providing. It’s clear that the long-term prospects of this site are limited; I decided to cancel the hosting plan.
At the same time, it’s true that AI can help people create content more easily. Whether it is subsequently consumed in a way that makes financial sense for the creator remains to be seen.
And the ability to rapidly create web apps could spur demand, as we discussed on this stretch’s podcast.
The second part of Bidzos’ statement discusses how AI tools can help people find available .com domains to register. I absolutely believe this is true.
