Gianluca Fiorelli: And talking about this, you already introduced the topic, which is something we will probably repeat a lot now. Internal links are the connectors, the axis, let’s call them so, that shape a content hub, a topical content hub.

So a hub would not be a hub without internal linking. And the classic structure is from the pillar to the sub-pillar and cluster contents, lateral between sub-pillars and cluster contents, and also the way back, from the clusters to the pillar.

But where I see many topical hubs failing is when people are not able to do correct lateral internal linking because maybe they are linking to content that is part of a topical hub but not strictly related to the specific sub-niche. For instance, a cluster content could be a landing page or a blog post.

Do you agree with me that this is usually the classic mistake people make when creating topical hub architectures?

Ryan Jones: I do; I absolutely agree with you on that. And I think this also connects to what we are seeing now and what we can do with things like the OpenAI API and things like that and getting into vector embeddings.

I have not done too much work around this. We have done a little bit of work internally and seen positive results. We know that internally linking where it makes sense for an algorithm can have positive results.

But going back to what I said before, and this links to the fact that no one, in my experience, knows your site or your topic. If you are a true expert in a particular topic, no one knows that better than you. So if you are looking at your site and reading through one of these sub-pieces, and you think an internal link laterally to a landing page or another content piece makes sense, then it probably does.

You can even see that if you do a bit of user testing on your site using something like Crazy Egg or something like that. If it makes sense to you, the chances are the algorithm is going to agree with you in some sense, even if you are not going into things like vector embeddings and saying we should place a link here because of that.

It really comes down to common sense and common practice. If you think it makes sense to do it, it probably does.

The downside is that you can go too far in the other direction. You might think, "Okay, we mentioned a keyword here, so we should link to that and that," and that is where people start creating messy content hubs and overlinking between them.

So there is a limit to how far you can go. But I still go back to that idea of topic ownership. If no one knows the topic better than you, then it makes sense to trust that judgment.

Gianluca Fiorelli: And another thing I want to ask you, related to topical hubs, is another mistake that I see and that I always talk against, which is considering topical hubs only within the constraints of informational content.

I think, and I do not know if you agree and if you have seen positive results, maybe doing the same thing I do, that topical hubs are not just about owning a topic on the informational side, but owning the search journey people want to follow around a topic.

So from very broad informational content to deeper levels. For instance, if we are creating a topical hub about sneakers or running shoes, you create content about how to choose the right sneakers, how to make them more comfortable after the first use, and so on.

Then you go deeper into very specific category pages of an e-commerce site, product pages, and so on. Or even having this kind of middle-of-the-funnel content, like the classic help pages about the product.

So, resolving issues of people who already had your sneakers or had other sneakers of another brand, and want to see if yours have the same problem. Do you think this is the correct way to create a topical map for conquering the so-called topical authority?

Ryan Jones: Yes, absolutely. And I think the way you described it is, in my way of looking at it, quite possibly the best way to do it.

We can have a whole separate discussion about what is happening to the top of the funnel in LLMs and Google’s AI Overview. I still think top-of-the-funnel content matters; it might just matter in a different way. We might not necessarily see clicks from it directly, but we can still see some clicks, and we still do from our own top-of-the-funnel content.

But we also need to make sure that we are the ones informing the discussion so that when LLMs go and do their grounding, we are present in that top-of-the-funnel search journey.

Without going too far into that discussion about LLMs and the whole death of top-of-the-funnel traffic thing, in terms of the actual journey, that is exactly where we want to be. We want to be known as the source that provides all the information someone needs when they are first looking into a topic. We want to make sure that we’re the source of information that people see when researching different tools or different websites to look at, or, like in your sneaker example, the different brands people could use and things like that.

And then, of course, we want to be the place where people go to make that purchase decision and choose to buy from us.

And absolutely, it also helps to have post-purchase content, like help-based content. Because even if we remove the entire SEO aspect from it entirely, it just works from a business standpoint for us to be the source of information across all aspects of the spectrum, from initial information right through to post-purchase, and making sure that people have a great experience with customer support if they need it.

Because that, at the end of the day, is what drives repeat business, and reduces churn if you are a software business. That is where we get into the whole business aspect of SEO. So even if you remove the SEO benefits from it as the first consideration, it still works from a business perspective.