In this guide, we’ll explore how you can put generative AI to work in your SEO strategy to help increase output.
The UK government recently announced a £100M taskforce to help boost safe AI adoption. With everyone talking about ChatGPT and similar AI systems, it’s no surprise that governments are starting to pay attention to how this plays out.
AI has the potential to upend a number of industries, but it could also help to transform the way we do a lot of marketing processes. Rather than thinking about how AI will replace SEOs and SEO copywriters, it’s better to think about how SEOs and copywriters can use these tools to 10x their marketing output.
What is generative AI?
Generative AI is a general term given to a field of computer science concerned with the processing of information by machines. You may have already encountered AI without even realising it. It typically includes an element of natural language processing and machine learning.
When we think about AI, thoughts often run to science fiction scenarios where machines are taking over. We’re not quite there yet, and AI should instead be seen as a tool at our disposal rather than a sentient entity.
Generative AI can also include things like words, images, audio and 3D models. An AI image recently won a prestigious photography award, which has generated mixed responses about the future of the medium.
From an SEO perspective, generative AI can be used to create things like website content, content briefs and even technical SEO documents. Let’s look at some of the ways we can use SEO in marketing.
How to use AI in your marketing strategy
The important thing to remember about AI is that it stands for artificial intelligence, and the artificial part is very important. It cannot replace human insight just yet, even if it does do a faster job or sorting information. But these are some of the ways that AI can help streamline the work of a marketer.
Generating content ideas
Churning out AI content and hoping it will connect with your audience/ solve a problem is a poor marketing strategy. Instead, think about ways you can streamline the content production process by coming up with content outlines using AI.
Copy.ai will allow you to create a blog brief from just a few prompts. Once you have your target keywords, enter your blog title and the tone you’re trying to achieve to land on a blog post outline that will shape your content.
You can then hand this brief to a copywriter, or handle this task yourself. Either way, you can quickly generate more content for your site with this handy tool.
Creating new images
Using AI to create images for your blog posts could help you to avoid the time-consuming process of searching royalty-free image websites for the perfect image. Tools like DALL.E will allow you to generate an image based on a prompt, which will enable you to create the perfect niche image to communicate your intentions with your content.
DALL.E is also capable of something called Outpainting, which allows you to extend an image in any direction. If you have found the perfect portrait image for your website, but you need it to be landscape, you can ask DALL.E to extend the image. The results are hit-and-miss but often very impressive.
Handle technical SEO and schema markup
Did you know ChatGPT can also generate things like XML sitemaps and robot.txt files. If you dread creating things like this, then ChatGPT can make light work of it using just a few simple prompts.
You can also use tools like wordlift.io to create the relevant schema required for the knowledge graph. This can be incredibly time consuming, but this tool makes light work of adding structured data to your website.
Using AI for redirects
When migrating a site, redirects are likely to cause more than a little bit of a headache. But with AI you can create a redirect map that cuts down time spent on this task by around 50%. Tools like getdirects.com will match URLs based on H1s, page titles and URLs, which is a large chunk of the work done for you.
It’s important to check all work before setting up redirects, as mistakes could cause a headache, but this tool is a great place to start. For anyone who thought AI could only help with churning out blog posts, it’s time to reconsider this stance.
Spot AI generated content
While AI has some useful applications, we don’t want to be using automated content without our knowledge. If you are concerned about AI generated content on your site (perhaps you have a large pool of contributors) then try tools like GPT Radar, Draft & Goal or Originality AI. As fast as AI is moving, so are the tools needed to detect when it has been used.
Best practices for adopting AI for your SEO content
While AI tools might be easily accessible and widely available, this doesn’t mean you should rush forth and get carried away with the possibilities just yet.
You need to confirm you have authorisation from decision makers in your company and ensure that you have weighed up the pros and cons of using AI web copy. Here are some best practice rules to consider when deciding if you should adopt AI for your SEO content production needs.
Do we really need AI?
Yes, it might be on trend at the moment, but do you really need to use it? Consider what business needs you are hoping to solve and ask if AI is the only way to achieve this. For some situations, AI could even make the task longer and more complicated.
The time taken to check the work and ensure it’s correct before it goes live should also be taken into consideration. In some situations, creating from scratch would be quicker than editing poorly generated content.
How much will it cost?
AI is not a free service, so you need to consider the real-terms cost of using it. Almost all AI tools have a subscription fee that is linked to the number of words generated per month. If you’re planning to lean on it heavily for content production, expect the cost to quickly escalate.
You also need to consider the time investment required to handle the AI tool. Someone needs to be in charge of managing and editing the output. Even researching the right prompts to use can be time consuming.
There is an incorrect assumption that AI is a free tool and that it will also free up time in your day. In reality, it’s a paid tool that also requires a time investment.
Create better prompts
You might not get the prompts right the first time, and this can lead to some frustrating generations that are either inaccurate or simply miss the mark. AI tools mirror what you put in, so if you don’t spend some time creating powerful prompts, you can expect to get poor generations in return.
Many AI tools offer feedback options that will allow you to reject some completions. You can then rephrase your prompt and try again, allowing the AI tool to learn from your feedback in the future.
Always check for plagiarism
Duplicate content and plagiarism are serious problems for SEOs. Whether a human is writing your content or an AI bot is writing your content, you need to be checking for plagiarism.
AI often lifts content from other sources, so don’t be surprised if you see long phrase matches in your AI generated content. This is one reason that generative AI should be used as a starting point rather than using it to create a finished product.
How can AI generated content go wrong?
The technology might be exciting and moving quickly, but it certainly isn’t without its flaws. Being aware of the limitations and potential issues with AI can help you to avoid some all too common pitfalls.
Remember that AI is currently under regulated, but this could all change in the future. If you choose to use AI for content creation, it’s important to keep documentation so you know what steps you need to take in the future if rules or guidelines change. Also, AI content is detectable with tools such as copyleaks.com. And, I believe Google is also trying to detect this too.
AI tools may be guilty of copyright infringement, but we don’t yet fully know the extent of this problem. Checking for plagiarism before you publish any piece of work is essential to help protect yourself. Editing AI generated content for clarity and originality is also a good way to ensure that you don’t get caught out.
There have been a few high profile cases of companies and individuals getting caught sharing content that has been ripped off from other sources.
- A substack newsletter was found to have plagiarised an article written by a human word for word.
- Stable Diffusion found itself in trouble after its AI generated images included the Getty Images logo, which indicates that their images were used to create the new images.
- Sites like CNET have been caught publishing AI generated content without proper checks.
At the moment, there is no obvious recourse of human writers who are plagiarised by AI content, but this could change in the future.
We also don’t know how human readers will react to AI generated content. It could increase mistrust of brands that are open about their AI generated content, or try to conceal their AI written content but get caught out another way.
What we do know is that readers generally don’t like to be pushed onto an AI chatbot when they want to speak to a human, so content generated by AI might elicit a similar reaction.
When an AI content generation tool doesn’t know the answer to something, it’s often prone to making up something entirely random. These are known as hallucinations, and they could be problematic if AI content production becomes more mainstream.
The content that is produced often sounds plausible, but it is completely fabricated, making it hard to detect unless you are thoroughly fact-checking your work. The issue here is that, once published, other AI generative tools might rely on this hallucination for research, which could mean that we double down on the false statements.
Before long, it will be difficult to know what is true and what has been fabricated by AI content tools.
Lack of originality and insight
If you are a thought-leader in your industry, people turn to you for genuine human insight. This cannot be replicated by an AI content creation tool, because AI is not capable of generating anything new. It only repackages what insight is already out there.
AI won’t create anything new or replace real human insight. And we know that Google and other search engines place a great deal of importance on the originality and insight offered by humans. Google’s E-E-A-T update placed experience, expertise, authoritativeness and trustworthiness at the heart of their operations. And at the moment, these are not qualities that AI can replicate.
The risk of harmful content
When AI is mining information with minimal filtering, there is the risk it could produce something that is harmful or offensive. Content could be considered racist or sexist, which highlights the importance of still having humans check and review everything before it goes live. When AI is pulling content for biassed sources, you can guarantee that the content it generates will also contain the same biases. Hitting publish on AI generated content without a thorough review and fact check could land you in trouble.
Won’t work forever
There are some companies that are jumping on the opportunity to quickly churn out lots of blog posts and content in the hope of getting one over on the search engines. This would likely be considered to be black hat SEO, and it won’t be long before search engines decide how they would like to deal with it.
AI generated content could be penalised in the future, which could place publishers at risk of falling rankings if they choose to move forward with an AI strategy without having a contingency plan in place. We can anticipate that Google and other search engines will improve their AI content detection tools and may penalise content produced by AI, particularly for Your Money Your Life (YMYL) niches.
The best generative AI tools to try
If you’re curious about how AI works and how you could use it to expand your content offering, here are a few different tools you should explore.
Perhaps the most well-known of them all, ChatGPT has already become synonymous with generative AI. It is built on the OpenAI system and is incredibly easy to use. The free version isn’t always available, but you can upgrade for $20 per month to access the tool at all times, and make the most of exclusive features. It’s worth noting that if you are getting something for free, you aren’t the customer, you’re the product. So at this stage, anyone using free AI tools should expect to be contributing to the product for free.
The image arm of OpenAI will allow you to generate images based on a text prompt, or upload and edit an image. You’ll need to purchase credits to use this tool, and there is no guarantee the image won’t pull from copyrighted sources.
Copy.ai is a generative AI tool aimed at marketers. It claims to be able to write blog posts 10x faster, write higher converting social media posts, and write more engaging emails. While this might be an excellent tool to speed up these processes, we’d still recommend thorough reviews to ensure the content is accurate.
Jasper is built on OpenAI and offers a wide range of templates to help you automate SEO and content production tasks. You can ask it to create product descriptions for sites like Amazon, or to create meta titles and descriptions based on keywords and a company or service overview. You can even ask it to outline and write an entire blog post based on a few prompts. As with all AI tools, the results are hit and miss.
AI might be changing the way we work, but it isn’t going to completely replace SEOs or content creators just yet. The landscape is constantly changing, so proceed with caution if you decide to add AI to your SEO toolkit.
It’s also important to make sure you are checking everything that your AI tool produces for plagiarism and accuracy. And finally, remember that humans should still be making the decisions, you can use AI to help inform your decision making, but it shouldn’t be making the big decisions for you.
I'm Paul Gordon an SEO freelancer who has spent the last 16 years helping businesses expand and grow through effective SEO strategies.