How to integrate generative AI in your SEO

As the initial buzz around AI dies down, the practical use of generative AI is now well-established in the content creation world. 

In theory, this creates a huge content marketing opportunity. However, as we are seeing, there are many pitfalls from an SEO perspective. 

This article outlines how we are primed to make such mistakes, ways Google has tackled this historically and how to safely integrate generative AI in your SEO and content marketing. 

Pitfalls and problems

Human beings are hardwired to seek the path of least resistance. 

Generative AI tools seemingly make it easy to write an expert-level article about any topic. This allows websites or authors to assume unwarranted expert status and pump out legions of articles on almost anything. 

This is the path of least resistance for content production, and on the surface, it seems an attractive proposition with some huge SEO opportunities.  

I have sat in marketing meetings with AI consultants advising this very strategy. One consultant called the approach “automate and dominate.” I was there to temper such blind enthusiasm and ensure no SEO mistakes were made. 

This creates a problem for consumers looking for the best information and for search engines who need to ensure that the best information is promoted accordingly. 

SEO history lesson

To look into the SEO future, we have to look at the well-documented SEO past. 

In the early days, around 1999-2000, it was easy to manipulate Google with links and content.

However, Google soon caught up and plugged the main problems of low-quality links and mass-produced, low-quality content with the Panda and Penguin algorithm updates. 

Today, it is easy to look at generative AI and believe these tools provide an opportunity to create lots of content quickly and grab a greater slice of organic search traffic. 

However, the battle-hardened Google of 2023 differs from the Google of old. 

High-quality and human-first content is already baked into the search algorithm, and recent updates have only served to confirm that. 

Google is ahead of the curve here, and we have already reviewed several sites that have seen recent dips in performance after helpful content updates, some with up to a 70% drop in organic traffic. 

There was nothing wrong with this site. It was not spammy, but it also did nothing unique. It was just the same content you could pick up from several sources. 

Reviewing the losses, it seems that rankings were lost to sites that were more expert or relevant for the topics. 

This is an existential threat to sites of this manner that rely on organic traffic but only pump out relatively generic content that could be found on many other sites. 

If SEO is central to your marketing strategy, then sensible integration of generative AI, rather than full automation, is the only sustainable strategy. 

What does Google want?

To understand what your content strategy should look like, we must consider what Google is trying to achieve and aim at that. 

Fortunately, Google communicates fairly clearly what they want within their documentation on creating helpful, reliable, people-first content. 

If you want to do a really deep dive, then take a look at the Search Quality Rater Guidelines (but at 176 pages is a somewhat time-consuming read). 

Fortunately, E-E-A-T summarizes the philosophy for the rest of us more concisely.

I don’t want to reinvent the wheel here, but the key points are as follows: 

• Experience: Personal experience should be clearly demonstrated in content.
• Expertise: Content should be based on an individual’s proven expertise.
• Authority: The author and website should be an authority on the topic.
• Trust: The author and website should be trustworthy. 

The key here is to factor this into your SEO goals. 

• Have content that is based on your real human expertise and experience.
• Ensure you have the technical aspects of your UX, website design and SEO in a good place. 

An AI analogy

The problem with SEO and generative AI reminds me of a scene in one of my favorite movies: Good Will Hunting. 

The protagonist, Will, is a genius who can consume, remember and regurgitate raw knowledge and data unlike anyone else alive – much like ChatGPT. 

Will knows everything to know, but it is all book smarts. He has no real-world experience. He has never lived, he has never loved, he has never lost someone close to him. He has never truly experienced any of what he knows. 

The AI content tools are a lot like this. They know everything about everything, they have crazy book smarts but no real experience, and they never will have. 

That is the opportunity where you can come in and create something unique by integrating these two components – your experience and expertise with the rapid output and breadth of knowledge provided by AI. 

Dig deeper: AI can’t write this: 10 ways to AI-proof your content for years to come

A strategy to safely use AI in your content production

The key here is to think of AI as an intelligent assistant rather than a complete means to an end. 

If you take the path-of-least resistance approach, as most will, your content will never satisfy Google’s criteria. Worse still, you will create nothing that Google can’t answer directly with their own Search Generative Experience (lose-lose). 

By combining your expertise and experience with generative AI, you can create content that is greater than the sum of its parts and almost certainly more significant than anything a generative AI can automate and produce alone. 

A simple SEO SWOT analysis that looks at your competitors will allow you to identify SEO opportunities, as most sites are just not doing this as well as they could yet. 

The benefits here extend beyond content creation. Whether you rely on SEO or PPC, or some other traffic generation, your engagement and conversion rates will rocket. 

Tasks where generative AI can be helpful

This is not to say that generative AI is not truly helpful. 

There are several tasks in the content ideation and creation process where generative AI tools like ChatGPT can radically help you speed up and improve your output. 

1. Ideation 

ChatGPT simplifies generating ideas and a simple prompt asking for topic ideas will generate several results. 

For example: “Can you suggest some article topics around generative AI and SEO?” 

This produced around 20 articles in 10 categories in around 30 seconds. Each of these could then be further explored (using your own expertise and experience). 

2. Research and information gathering

AI can be used to quickly sift through vast amounts of data or provide summaries of detailed articles or scientific studies. This can help you develop deep, well-researched articles in a short amount of time. 

You can also use ChatGPT to cross-check facts and provide sources to ensure the information shared is trustworthy. 

It is always worth double-checking before you publish, but even with this step, you will speed up your workflow and cover more ground. 

3. Content drafting

With a topic and a rough outline, generative AI can quickly create drafts for you to polish and refine. You can also create variants in different tones, styles or from different perspectives to pick the very best approach. 

4. Editing and proofing

AI can check your grammar, review content readability and suggest improvements in flow and structure. 

5. SEO

AI can help simplify the optimization of an article by suggesting keywords and optimizing the article and metadata. 

6. E-E-A-T 

AI tools like ChatGPT can also review your content from the perspective of E-E-A-T criteria and suggest improvements to ensure you hit all of these important points for modern SEO content. 

Tip: I would generally recommend tackling each element of the E-E-A-T criteria separately here, as you can get some whacky results trying to get feedback here. 

7. A/B testing

AI makes it easy to create multiple versions of content headlines, calls-to-action or even an entire content piece and then use these to split tests for engagement. 

How to review your content 

Google has outlined what they are looking for with E-E-A-T and also provides a series of questions you can use to review your content at Google Search Central.

Google certainly provides a comprehensive set of questions to review content. Still, for your average content writer or marketing team, this is almost an insane level of overkill that will tie you up in knots. 

To make this more actionable, I have created a simple set of 10 questions that covers the author, content, website and E-E-A-T and helps you ensure your content is something that Google will want to surface in the search results. 

Author

• Is it clear who authored the content?
• Is there detailed author information available?

Content 

• Is the content original, useful and substantive? 
• Is the content different or substantially better than other pieces out there?
• Does the content provide a new viewpoint or way to view the topic?

Website

• Is the publishing website credible with a good reputation?

Experience

• Does the author have first-hand experience on this topic?

Expertise

• Does the author have demonstrable expertise on the subject?

Authority

• Is the author (and or website) a known expert and source for this subject?

Trust

• Is the page accurate and the facts reliable?

You can also use tools like ChatGPT to help you answer these questions and find out more about the author, review the content, review competing content and much more. 

You just have to get creative in your approach, ask these tools questions and review the responses. 

AI in SEO: Intelligent assistance 

The obvious pitfall with generative AI and SEO is falling into the path-of-least-resistance trap. 

SEO history has shown us that these low-value approaches, while you may generate some small, transient wins, won’t win the more significant, long-term battles. 

Instead of looking to automate and dominate, use AI tools as an intelligent assistant rather than a complete means to an end. 

Automated content will never satisfy Google’s ever more robust criteria. Worse still, you will create nothing that Google can’t answer directly with their own Search Generative Experience, making your content irrelevant.  

By combining your expertise and experience with generative AI, you can create content greater than the sum of its parts and almost certainly greater than anything a generative AI can produce alone. 

The benefits here extend beyond content creation, and whether you rely on SEO or PPC or something other traffic generation, your engagement and conversion rates will skyrocket.