Hub-and-spoke content: How we increased organic traffic by 300% in 6 months

When I joined Factorial, an all-in-one HR software for SMBs, it was a 75-person Series A occupying one floor of a nondescript Barcelona office building. It had seen some early success in Spain and was working toward opening other markets in Europe, South America, and (of course!) North America.

As the first and only marketer dedicated to the United States, I was responsible for lifting the marketing function off the ground and getting us the leads we needed to raise a North America-centric Series B. But resources were tight; we had to be scrappy. If you’re looking for a cost-efficient initiative that comes with a big, big return on investment? Organic traffic is the way to go.

There was only one problem: When I started, traffic to the US site (factorialhr.com) was completely flat. At first, we struggled to find a way to move the needle. But by zeroing in on highly relevant topic clusters, I worked with my team to 3x organic traffic in 6 months — getting Factorial the traction it needed in the US market for down-funnel lead gen. Eventually, I would ghostwrite the announcement for Factorial’s $80 million Series B.

Here’s the story of how we did it.

The problem: Flat traffic and slow lead gen

Although Factorial was publishing content regularly before I joined, that content wasn’t bringing the kind of results they needed in terms of either traffic or leads. With a bit of poking around, we found the major problems:

No authority

Google evaluates the quality of content (and increases the visibility of quality content) through the infamous EEAT criteria: Experience, Expertise, Authoritativeness, and Trustworthiness.

Factorial was publishing on a broad variety of themes. While such breadth may seem like a good thing, it can actually undermines SEO strategy because there was not enough content around one theme to establish Factorial as an expert in any one arena. If we wanted to be seen as experts, we needed to focus content production on fewer content verticals rather than casting such a wide net.

Poor targeting

Of course, traffic is not an end in itself. Good content converts traffic into leads and the best content converts traffic into customers.

While deep-diving into Factorial’s website structure, I realized that a lot of the high-traffic pages were not aligned with Factorial’s core audience: HR professionals at SMBs. Our content strategy would need to niche down into more HR-specific for small businesses in order to address lead gen alongside traffic.

Creating a trove of audience-specific content would also help to establish us as credible authorities in the space, in the eyes of both consumers and search engines. This would make it easier to rank for these terms, creating a virtuous cycle that would drive more converting traffic to the website.

The solution: Hub-and-spoke content structure

Working alongside my wonderful Content Manager, Kat Valier, we decided to implement a Hub and Spoke content structure, building content clusters around core topics. Having content clusters around thematically and semantically linked content helps boost your site’s credibility.

These clusters rely on highly-directed keyword research and robust interlinking structure. Making sure your content is linked to other relevant content on the site is one of the best ways to establish your site as credible in its space. Here’s how we got started.

Step 1: Create your dream content structure

The first thing you need to do is dream big. What would a map of your ideal site structure look like? What topics would you cover and how would they relate to each other? This is an exercise you absolutely need a visual aid for. When I first created the structure for Factorial, I used one of Whimsical’s mind map templates, but you could do the same thing with Miro or Lucidchart. (Not to keep listing diagramming tools, but for the sake of creating shareable images for this portfolio post, I used Excalidraw.)

For us, the central keyword we wanted to rank for at Factorial, an HR software company, was “HR software.” That means we put “HR software” is at the center of our content structure because it is the word that we want the homepage, the center of our site, to rank for. If you’re thinking to yourself, “But isn’t ‘HR software’ a bit broad?,” you’re a great marketer!

It’s a good idea to target the highest difficulty words on the homepage because the homepage of any website gets the most SEO “juice.” “HR software” is a very important, high-intention keyword, in that if you are searching for “HR software,” you are investigating, probably for purchase, HR software. In order to tailor this content more specifically to our SMB audience, we will pursue longer tail keywords on other pages. So here it is — the central hub of our hub-and-spoke structure.

Next, we need to build out the content cluster, semantically supporting that keyword with other related keywords.

Factorial’s all-in-one software had a lot of HR functionalities but our time tracking software and PTO tracking software were some of the most beloved features. Because these are fairly high-intention keywords — if you are searching for these terms, you are almost certainly interested in acquiring such a software — these pages are our public-facing product pages.

Unlike blog posts which are more informational, public pages are a direct pitch of your software and geared toward conversion rather than awareness. These two public pages remain some of Factorial’s best performing content.

Next, we decided to support those public pages with blog posts. Every public page, we determined, would have supporting blog vertical. The lowest-funnel, highest-conversion words live on the public pages while the higher-funnel, more informational keywords live on the blog.

People who land on the blog might not necessarily be interested in purchasing software right now (if they are interested in purchasing software right now, they should be on the public pages!) — but they should be in HR roles at SMBs, maybe troubleshooting issues with time tracking and PTO tracking.

So we want to semantically support the “Time tracking software” public page with a blog post that has broader, more informational keyword: “Employee time tracking.” This blog post will become a hub for the content cluster on the blog, in that it will link out to every other related post and every other related post will link back to it — as well as the “Time tracking software” hub.

To avoid getting too far in the weeds here, you continue building out this dream structure for the other dream keywords you want — thinking of them in content clusters and keeping high-intention converting words on the public pages and more informational keywords on the blog. Of course, keep in mind that best keyword strategy will solve problems that your customers face. But you can hammer down the broad shape of your site structure before getting into the nitty gritty of long-tail keywords.

Step 2: Map existing content using this structure

If you are building a website from scratch, you could use this map to guide your content creation and be well on your way to content nirvana. If, however, you have a legacy website that has seen many eras of content and growth managers, you will want to make sure that you are accounting for existing content with this content cluster.

Intentional internal linking is absolutely key for the hub-and-spoke structure and linking to new content from existing content (especially if that existing content is high-performing) is one of the best ways to get it recognized by search engines. Every time you create a new article, it needs to be included in this map and to have inlinks from and outlinkes to more established older articles. Using this map as a source of truth will also ensure that you don’t waste resources creating duplicate content and give you a clear sense of the optimization opportunities.

Factorial’s US site had about 300 (!) existing articles when I came in — many of which I had written myself as a freelancer under the direction of the previous content manager — and there was a lot of housekeeping to do in terms of content cannibalization and creating a coherent internal linking structure.

Using old faithfuls Analytics, Google Search Console and SEMrush, we were able to create a linked-out map of all the content in clusters, which helped give us clarity on where to move forward.

Step 3: Creating new content

Once you have grounded your dream content map in your current content reality, you are on your way to Steps 3 and 4: Creating new content and optimizing old content. It is a bit misleading to sequence these steps when really, they will need to happen in tandem, but one can only ever do one thing at a time.

At Factorial, we focused on two content clusters a month, both in terms of new content production and optimizations. In terms of creating new content, we were commissioning about twelve new articles a month from a wonderful freelancer, Cat Symonds, about 6 articles for each content cluster. Meanwhile, Kat Valier, our Content Manager, would optimize existing articles in those clusters.

Each month we tried to concentrate on one content cluster that was high-performing (because time investments for these clusters would show the quickest results) and one that was not performing so well (because we wanted to build these up for the future).

Our approach to each cluster would vary, depending on the existing content that had been produced and the keyword opportunities in the cluster. Sometimes we needed to build a hub post to unify several articles that had more long-tail keywords. Sometimes we needed to build spoke pages in order to support a broader hub page and nail more long-tail keywords.

We used customer interviews, both from the Sales cycles and from Customer Success, in order to ground our content strategy in content that would be truly useless for our audience.

Step 4: Optimize

Even as we were commissioning new articles, Kat was dusting off older to content to help it performa better, sometimes removing content that was no longer relevant or merging articles that had duplicate content and were cannibalizing each other. It was important, for a site as sprawling as Factorial’s, to be ruthless with content that wasn’t performing.

One of the most important parts of Kat’s work was looking after the internal linking structure: making sure that new content linked to all the previous content produced in the cluster, and vice versa. It isn’t enough just to drop a link in the blog post: This internal linking structure is most effective when the link is under a heading that is related to the keyword. It is also important to use semantic variations of the blog post’s keyword as anchor text for the link.

I frankly attribute the success of this campaign to the publication cadence we were able to build, as well as the internal linking structure.

The future of SEO

SEO is a bit out of vogue.

Especially this kind of SEO, which may strike folks as a bit mercenary. I’m aware of that! With the rise of ChatGPT and other AI tools, this kind of SEO is probably going extinct — primarily because search engines will be less likely to send people to website for information and more likely to scrape results pages and summarize them.

Although this strategy worked exceptionally well during the time that we implemented it, I’m not sure that it would be worth the investment to recreate today. My view of content creation now focuses more on the kind of unique, creative content beyond the scope of ChatGPT over informational content that will very soon cease to be a source of traffic.

That’s the kind of work we did at Garden — another tale for another time. In the meantime, please feel free to get in touch or check out other projects.