Topical Authority is still a mystical concept, partly because we don't have a good way to measure it. In this post, I want to offer a method to quantify Topic Authority.
I’ve used your way of measuring topical authority in the past, and I can’t think of a more practical proxy for doing it.
However, it would seem to me that your model would predict a great degree of topical authority for sites like Forbes, where I think it would be fair to say that other signals (“domain authority” based in large on “non-topical” backlinks if measured by any given topic it ranks for) better predict its rankings.
So my question is really if you agree with that, and if so, whether you think it’s possible to get a “cleaner” measure of topical authority that is separate from overall site strength?
Publishing a tool in collab with airops next week that will do a more granular job. I will say, though, that topical authority is very influenced by domain authority (backlinks) and more of a compound metrics driven by many small things.
I’d be excited to see what you share, but I do wonder if any model that is based on ahrefs rankings and keyword search volumes, will produce similar results to one based on measuring visibility on LLMs/AIO/AI Mode (not that we currently have good data on any of that).
Not trying to be contrarian here though.
The point being, i don’t assume that links will carry the same weight for LLMs beyond a certain baseline, and that after that vector relevance might be more important. If so, decoupling topical authority from links, or at least analyzing it as a composite metric by breaking it into several independent metrics, might be key to measure topical authority on LLMs.
Perhaps not helpful, but it’s food for thought. :)
(heck, even my Next.js “practice” site gets clicks from LLMs and I’ve done about zero actual SEO for it and only earned a few quality links naturally)
I’ve used your way of measuring topical authority in the past, and I can’t think of a more practical proxy for doing it.
However, it would seem to me that your model would predict a great degree of topical authority for sites like Forbes, where I think it would be fair to say that other signals (“domain authority” based in large on “non-topical” backlinks if measured by any given topic it ranks for) better predict its rankings.
So my question is really if you agree with that, and if so, whether you think it’s possible to get a “cleaner” measure of topical authority that is separate from overall site strength?
Publishing a tool in collab with airops next week that will do a more granular job. I will say, though, that topical authority is very influenced by domain authority (backlinks) and more of a compound metrics driven by many small things.
I’d be excited to see what you share, but I do wonder if any model that is based on ahrefs rankings and keyword search volumes, will produce similar results to one based on measuring visibility on LLMs/AIO/AI Mode (not that we currently have good data on any of that).
Not trying to be contrarian here though.
The point being, i don’t assume that links will carry the same weight for LLMs beyond a certain baseline, and that after that vector relevance might be more important. If so, decoupling topical authority from links, or at least analyzing it as a composite metric by breaking it into several independent metrics, might be key to measure topical authority on LLMs.
Perhaps not helpful, but it’s food for thought. :)
(heck, even my Next.js “practice” site gets clicks from LLMs and I’ve done about zero actual SEO for it and only earned a few quality links naturally)
Agree that links matter much less for llms, probably more passively
After reading your stuff, I'm instantly reminded of how little I know.
don’t forget how MUCH you already know, too
Nice piece!
thank you!