Back in September 2023, I wrote an article ‘An SEO Experiment on Semantic Strategies, Topical Authority and AI-Driven Content‘. The project aimed to enhance the digital presence of this fine website by integrating Semantic SEO and Topical SEO Maps, using advanced AI for content structuring, and all this being supplemented by human editing.
Collaboration with an SEO content agency ensures adherence to the brand while handling the bulk of the task. The three-month experiment includes:
Solid Interlinking: Strengthening the website’s structure through internal links to improve user navigation and SEO.
AI-Generated Content Outlines: Employing AI to create initial content frameworks based on semantic relevance and authority, providing a basis for comprehensive articles and blogs.
100% Human Editing: Fine-tuning content with human editors to guarantee quality and relevance.
Note: The human editing aspect is the real kicker here, of course, and overlooked by many of the AI content punters out there, who literally let AI do everything. Human editing is a necessity with AI content, as many have found. Google knows what AI content looks like because it knows what human content looks like. E-E-A-T is vital, and that will always come from an expert with authority, and actual knowledge they are sharing. An AI can be programmed to create content as a persona, sure, but the extra steps involved are often too much in the workflow for the content barons of the world. They are looking to produce as much content with as little involvement as possible. We’ll see these sites disappearing from view, no doubt, in a big “AI content update” from Google in the near future.
Success was measured by:
Increased Ranking for Keywords: Aiming for top 10 rankings, focusing on high-value, long-tail keywords to drive significant traffic.
Boosted Monthly Traffic: Aiming for a noticeable increase in organic website visitors.
Expanded Geographical Reach: Targeting a broader, global audience while also concentrating on specific UK areas like Manchester.
Ranking a Specific Keyword: Aiming to place a two-word, brand-related keyword within Google’s top 10.
SEO Experiment – Nuts & Bolts
If you have ever run an experiment with SEO, you know that it’s very easy to see the early wins and dial in on what’s working. Wherever there are patterns in Analytics and GSC (when you’re doing something different from usual), you can create on-the-fly experiments and side missions. This was the case for the most part – finding content that resounded and ranked, then creating more content like it.
So, what you’re seeing is an experiment that evolved over the course of the project. There was a plan (The Semantic SEO Topical Map) that was originally created, which had potential content titles and keywords, with all the potential internal links… and then there was ad-hoc content that was created based on early results. Overall, this is a very viable strategy for most people, and is industry agnostic. Anyone can do this.
On top of all this, there was content that could be upgraded and updated, and this was successful too. This shows Google that you are committed to your content and keen to revitalise good content. It is always possible to get better rankings just by updating a six month old article, and adding new data & images. Question-focused content also proved great to revitalise.
So, with all of this in mind, here are the results!
SEO Experiment Results
In the below screenshots, you can see the point when we began the experiment, and the current results in the SERPs. These are from Ahrefs data and SEMRush data, as well as Google Analytics and Google Search Console.
SEO Experiment Start Date: 19th September 2023
Data Compiled: 17th January 2024 (120 days elapsed)
Over this period of time, 32 articles were published/updated.
Google Search Console
The comparison below is the previous 120 days (22nd May to 19th September) and the period of the experiment (19th September 2023 to 17th January 2024).
Here’s those splits:
Top 20 keywords
SEO Experiment Results Analysis
It’s plain to see from the amount of traffic and number of new keywords appearing in the top 20 that the experiment proved to be extremely worthwhile, and positive. I also received a few contact form entries, and these were a pleasant surprise. Nothing was set up to be a call to action or a funnel in any way, and the focus was on content alone. To me, this suggests there’s much more can be done to turn this into top of funnel education content, with a massive CTA at the end.
The content itself stopped being published in December, just before the Christmas madness, and very little has been added since. It’s great to see that this content is still working hard, however.
On top of all this, the Google Images rankings have improved for some extremely good keywords. For example: “aims grid“
This is very useful. For anyone searching and wanting a visual representation, the ranking is there in the images. This kind of ranking is particularly useful for educational stuff I talk about (as they often include diagrams). There’s a little nod to the site in these images too (bottom right in the below image).
Although the traffic is not earth-shatteringly good, and my server is not buckling under the weight – this is mostly due to the content we chose for the blog. The emphasis was on improving rankings for some technical marketing stuff and for ‘me’ personally (“Paul Morris marketing” and “Paul Morris digital marketing” for example). it’s important to note that all of this was foundation SEO.
The next step would absolutely to be to cast the net wider and attack keywords that are broader in terms of marketing and digital marketing. Older content would also be worked on and brought up to date, such as AIMS Grid and other helpful content.
Does it work?
As a strategy, this works. It is all about creating valuable content that has a unique expert voice. It ticks all the boxes for Google in terms of E-E-A-T (Experience, Expertise, Authoritativeness, and Trustworthiness), and is useful to the reader (hence the contact form entries!). For social, this content can be condensed and repurposed, too.
Will I use it in future? Potentially, yes. If I decide to keep updating the content on my website, and need a full strategy built around it to target specific keywords – absolutely yes. If I were creating a new site and wanted a 3, 6 and 12 month plan for SEO, there is little better than an SEO topical map and a semantic SEO strategy.
Again, though, I would strongly suggest a human editing aspect to any work that involves AI. We’re not there yet, folks, and AI still hallucinates and adds its own facts here and there. The last thing you want to do is mislead people and get a strike against your site for misinformation. Whatever tools you are using, and whatever agency you are working with – ensure the human factor is paramount.