What’s Next For Aged Domains In Our Recursive AI Hellscape?
This was originally published in my newsletter, Rank Theory.
Subscribe to never miss a post:
I want to talk about something I’m just getting into, how I got there, and what I’m doing… I think you’re gonna find it interesting.
The quick summary of this is: can we leverage aged domains for traffic by building them for LLMs specifically?
Like everyone else in the entire SEO industry, I’ve been paying attention to the increasing fascination with, and the slow bleeding of AI’s impact on organic traffic.
There’s one idea I keep coming back to like the dessert table at Golden Corral and also I hate myself…
Every brand with a capable marketing team is going to have two main sites:
One for people to buy their shit, that plays by Google’s best practices and doesn’t abuse any particular tactic too hard, and
One for LLMs that still mostly adheres to SEO best practices but pushes the envelope HARD on content.
I’m gonna break down what I think that looks like and how it might play out (because YES, this is the theory part of Rank Theory, and something I’m just beginning to fuck with it myself).
The Content LLMs Want Is Not The Content That Ranks In Traditional Search
My bud Travis recently posted a pretty juicy thread after he attended a GEO conference (I still hate that term, LLMs are not generative engines wtf is even that, NO ONE CALLS THEM GENERATIVE ENGINES).
For me, this particular tweet hits the heart of what I’ve been thinking about, an extension of the idea that LLMs process and rank content very differently from how Google’s algo processes and ranks it (probably? eventually??):
If you’re creating content for a topic, play with making different “entrances” to the same thing.
Different pages that are in different tones and different styles and different formats.
Why?
Because the queries are not keywords. They are specific and contextual to the individual.
You have to appeal to multiple individuals.
(but a warning, don’t kill your organic SEO at the same time with keyword cannibalization)
That last parenthetical statement is the one I want to really drill down on:
If you’re trying to rank a site in Google, you have one page about a topic where you optimize all your internal links, build all your niche-relevant links to, and whatever you do do not cannibalize your own content!!!
But if you have a site that you’re specifically trying to get mentioned by LLMs, you don’t really care about cannibalizing your content. If you’ve got a site that an LLM trusts, with a ton of niche relevancy already, it makes sense to me that you’d want a thousand chances at bat with different nuance than putting all your precious eggs into one rickety ole Google basket.
So how do you resolve this tension between a site with Google best practices and a site where you’re catering to our new AI overlords?
You Gotta Secondary Site That Shit
In the before times you’d grab an expiring domain with juice, put up the absolute worst spun content your $49/mo tool could give you, and build links to your network of money sites for da links.
Then people were like “wait I can just take this New Mexican Diner website that just expired and build a kitchen affiliate site on it and make $17k/mo with amazon!”
And that’s where we found value in aged domains for a LONG time–basically until the HCU.
Now all the best aged domains go to iGaming (even Odys, the one-time-champion of building aged domains in the pre-HCU era invested heavily in the iGaming niche):
(though they still sell regular aged domains as before).
But what’s next for GOOD aged domains?
Here’s my first thought leader-y idea:
I think they are going to be used as secondary sites for established brands.
Of course there is some overlap between what ranks traditionally and what “ranks” (get mentioned heavily) in an LLM. A ton of overlap–to the point of the common advice, at least as far as I’ve seen, being “if you want love from LLMs just keep doing good SEO and ranking well.”
But there are going to be really obvious differences.
If an LLM knows and trusts your site, in theory you can be a lot looser with things like content cannibalization. As mentioned in the tweet referenced above, there could be a lot of value in having different pieces of content with largely the same message, just framed differently.
So you’ve got your main site talking about your main product and you’re paying your tribute to the God of Core Algorithm Updates like a good little webmaster, but you’ve ALSO got this other site… aged, trusted, a known quantity to the LLMs (maybe it was used in the training data, which won’t help you directly until the next model update or however that works, but can still send relevant traffic your way)–whatever the case may be. You hit it with hella pieces of content across a multitude of highly specific styles/contexts/author personas, and you stack. those. mentions.
Is it worth it?
Will it work?
IDK, but it makes sense right?
And Google’s AIO is stealing all your clicks despite them lying about it right to your face.
AND, AIs really love to send traffic to all that sweet, sweet informational content everyone stole and trained their models on.
AND ALSO, according to Exploding Topics (owned by Semrush):
On average, a website visitor from ChatGPT is 4.4 times more likely to convert than the average visitor from traditional organic search.
The last piece of this thread comes from Duane Forrester, a smart SEO who wrote an interesting piece about how to approach structuring sites and their content when trying to suck up to LLMs (for the clicks):
The New SEO Work: What Actually Drives Results Now
Real SEO today revolves around structured, retrievable, semantically rich content:
Semantic Chunking
Creating content structured into clearly defined, self-contained chunks optimized for LLMs (Large Language Models).
Vector Modeling & Embeddings
Placing content into semantic clusters inside vector databases, ensuring each piece of content is closely aligned with user intent and query vectors.
Trust, Signal Engineering
Implementing structured citations, schema markup, clear attribution, and credibility signals that AI-driven models trust enough to cite explicitly.
Retrieval Simulation & Prediction
Using tools such as RankBee, SERPRecon, and Waikay.io to actively simulate how your content surfaces within AI, driven answers.
RRF Tuning & Model Optimization
Fine-tuning content performance across generative models like Perplexity, Gemini, ChatGPT, ensuring maximum retrievability in various conversational contexts.
Zero-Click Optimization
Optimizing content not just for clicks but to be featured directly in generative AI responses.
To recap:
- writing for LLMs requires a different approach to writing for Google’s algo
- it’s worth it because traffic from LLMs converts pretty well
- you don’t want to write purely for LLMs on your main site and risk Google’s wrath
- use a secondary site with trust and authority already established (the value of an aged domain) to run this strategy.
What Are The Implications For The Aged Domain Market
I think I’m early to this way of thinking, but it’s not like it’s a huge leap to get here, I’m just thinking about domains way more than your average SEO.
Eventually I believe many in the industry will realize the potential of the secondary-site-targeting-LLMs as an attractive strategy and GOOD, POWERFUL aged domains will become even more expensive (and lordy, they are so fucking expensive currently).
They are like rare-earth elements: there’s a limited supply, and you can create insanely valuable things from them. My best guess is another year or so before this becomes a top-3 reason people are acquiring these names.
I might write about this more in the future, but if you go back to 2019 – 2023-ish you’ll see (probably gotta hit archive.org) that most of the big aged domains with juice sales during those years had sites built on them/were 301’d to other sites.
If you look at the big aged domain sales from 2024 they are LARGELY linking to/becoming online casino affiliates.
I think in 2026 you’re going to see these domain rebuilds becoming secondary sites for LLMs.
Here’s my second Thought Leader-y idea:
I’ve always thought it could be worthwhile (especially at the sky-high prices of the last year or so) to try and find pre-expired aged domains–just parking it in a niche and searching for “dead site” signals, check against SEO value, and conducting outreach to try and land a site for like 1/10th the price you’d find on GoDaddy auctions.
I never quite had a lucrative and compelling enough reason to do this (it’d represent A LOT of work and I am out of the affiliate game), but I always thought it was the smart way to spend $800/mo on a VA + however much it’d cost to run cold outreach campaigns.
I *still* think that’s a good idea, but I think there’s a better way:
Acquire smaller sites in your space and turn them into LLM satellite sites.
The downsides here are:
- you could kill their Google rankings
- this is not a cheap strategy
But the upsides are pretty great:
- established sites with authority
- never dropped (if you care about that, I don’t)
- comes with content + rankings
- ready for more content ASAP
I’ve already mentioned my Acquisition SEO service, and I think this represents another strong case for this strategy, only instead of acquiring a site to take over and push to your product (Backlinko >> Semrush), or to put all the content onto your main site and 301 everything, you start testing shit for LLMs and hitting it with a ton of human-optimized AI content (which is apparently the gold standard of content ranking these days PROBABLY because the fact that this is what like 85% of sites are publishing anyway) and try and get your LLM-traffic way up.
It’s part of a balanced breakfast:
- acquire sites to 301 to your own
- acquire sites to build out and push to your brand
- acquire sites to turn into huge LLM libraries of info content from a ton of different angles, for different personas, and push your product
- (to say nothing of acquiring subreddits, youtube channels, email lists, social accounts, forums/communities, etc to prop all this up.
But, you know…
I’m a hammer with an owned-media-is-the-only-nail shaped head, so of course that’s where my mind goes.
And before anyone absolutely roasts me like chestnuts on an open fire, just know that I’m a sexy SEO expert, not a sexy LLM researcher so some of my assumptions about how LLMs work and how to make them rank you may be… wrong insane still evolving.
If you found these ideas interesting and want to work together, just grab a spot on my calendar here and we’ll see if we’re a good fit.