In the last few years, AI has gone from futuristic to reality in many fields of our lives. SEO is no exception. In spite of lively discussions over Google’s reaction to generated content, more and more people are testing and trying to adapt AI to their niches. So should or shouldn’t you add AI content creation tools to your SEO strategy? Let’s dive right in!
Research methodology
We asked 50 SEO specialists (among them Matt Diggity, Jason Barnard, Lidia Infante, Shane Barker, Ann Smarty, Andrea Volpini, and Adriana Stein) in different fields to go through the online survey.
There were 6 voting questions (from the type of content they generate with AI-powered writing tools to their favorite software for this aim) and 3 open questions about their workflow, experience, and predictions about AI in SEO.
We summed up the votes and created infographics based on them. You can view the raw data in this Google Sheet. Our study revealed some very interesting (and useful) insights. So now it’s time to share what we discovered.
Our summary findings are as follows:
- 70% of SEO experts have positive experience with AI-generated content but are still sure this content isn’t enough to get to the top of the SERP without human touch.
- 64% of voters use AI tools for generating blog posts and only 20% prefer to use them for generating images.
- 68% of SEO professionals believe that the age of websites doesn’t matter for generating AI content.
- 72% of SEO experts use Jasper in their SEO content routine.
Age and favored types of websites for AI-generated content
Almost 70% of SEO specialists are sure that the age of websites doesn’t matter for automate content creation but you should always keep in mind that quality content is still the king when it comes to SEO.
You need to be picky with niches where AI can add more “flavor” to paragraphs than mistakes. Mostspecialists use artificial intelligence in SEO to write,edit and summarize posts for blogs. At the same time, the quality and use of AI for e-commerce, social media websites, and forums are still low.
Due to the niche of the website (adult eComm) the test was unsuccessful as there were strict restrictions on what they could and couldn’t write about. As a result, we stopped using it. However, I would be willing to test it again, with updates like “helpful content updates” and related broad core algorithms.
I’ve noticed that some writers use AI to write product reviews and product features, but in most cases, they are wrong (i.e. measurements, etc). We don’t use AI tools except for our content briefs.
Type of content usually generated by AI SEO software
As for the type of content, opinions differ: some use AI for generating long-form content and others prefer it only when lacking inspiration for briefs or short-form content.
In terms of SEO, it could be a great help for meta descriptions and perhaps for adding smaller paragraphs.
I use AI tools for content briefs but I write the content myself.
My findings are that AI is great for producing short-form copy and automating h1 tags, meta titles, and descriptions but with long forms such as blogs, it doesn’t always work well because again searcher’s intent will need to be considered. It’s not something we can rely on without human input or sense-checking.
Image creation is a top use for AI-generated content, as well as using the text-focused tool for outlines and ideation. None I’d used can be trusted or used without human supervision and editing — but they’re great for doing what machines do best: collecting a bunch of data and formatting it the way you want.
AI content will be good for PBN’s pages or less important blog pages.
Ideas for implementing AI tools in SEO workflow
We use a G.E.O. method to get the most out of AI. G= Guide the tools with a preconstructed outline based on the top competitors. E= Edit the content for grammar and factual issues. O= Optimize entity frequency when you get to page 2.
We use AI tools to generate headers and title tags, having first established a set of semantically related keywords which we wanted to target. Using the AI-generated versions of copy, we further UX-tested these in a variety of ways. Firstly, using tools such as Crazy Egg to establish on-site engagement with the AI-generated content, and then using the same tool to UX test the different versions. We also rolled these out to different audiences in our database via email campaigns, and across various social platforms. The results were interesting… AI-generated content was helpful in producing multiple versions of content that helped us to test quickly, responsively, and at scale.
It is a bespoke experience where designers, ML experts, and content writers have to find their own magic formula.
We often work with content for our additional tools, such as the traffic checker and others. Opening up the functionality to users on this type of content is extremely useful when utilizing GPT. By uploading screenshots of the functionality and extracting a well-crafted description from it, and with some minor improvements, you can create a valuable content block for your webpage super fast. This method can be applied to various tasks requiring image description.
Еvaluation of the experience with AI content generators
70% of SEO experts are satisfied with the use of AI-powered writing assistant tools. It helps to think bigger and write faster with a creative AI idea generator-partner by their side. However, all of them are sure that it always should be double-checked by humans. So it’s better to test what is less money and time-consuming in your case.
Some experts have already implemented it in their workflow and they are content:
If you do lazy AI content generation (submit ideas, publish content) then the results are often poor for users and the business. However, if you take the proper time to treat AI like a tool (research, outline, submit an idea, curate/edit copy, publish content) then the results will be fantastic for everyone.
I decided to test it to save money. One content writer can produce more content using Jasper/Rytr/etc, in comparison with the classic method. I have the experience when proofreaders were checking our AI content, but they didn’t know it was one. So, they didn’t even guess it was AI. And ofc, that way of creating content can get the same high positions in search as classic content. About combining: if the content really needs human research and involvement, I delegate that task to professional writers. If it can be generated – it should be generated. But anyway, the content maker must do it and if needed, make some corrections.
If you’re looking to level up your content game, it might be the time to consider using AI. It’s cut my writing by 20% now that I’ve learned how to use these tools.
Some of them still have ambiguous feelings:
We all want to save money on content. For some of us, it’s becoming more expensive than links. However, for most of my projects, I cannot do that at scale because either the quality is awful or the cost becomes too high for this type of content. Nevertheless, experiments and some outputs are not only fun but useful. It is good for news, stacking, local, and easy niches. I use Grammarly to check the content. And use content managers a lot 🙂
AI tools can basically be a personal search engine that provides you with whatever you ask for. AI tools are like SEO tools; they provide results and insights, but without knowing, those things mean nothing but just results.
AI is able to create high-quality content right away. But since I do have my personal style, it gets messy when the AI takes over. So I have to instate that by adjusting the texts accordingly. Since “re-writing” now works quite well for me, I’m testing these tools in the content creation process for clients.
And some of them are disappointed with AI content generators at the moment.
When we compare the time it takes to create content in a manner that provides true business growth, we’ve found that no time was saved because either way content needs to be heavily edited by a person who has a deep understanding of the SEO strategy and business strategy as a whole. When we combine the time factor with the increased risk of issues with Google itself, we have opted to avoid AI tools. So as of yet, our vote has remained as a “no” to using AI tools for SEO.
Awful. Read terribly to human eyes and was deemed inexcusably bad for CX, so was aborted before it really hit any SEO level!
The biggest problems with AI content generation tools
The biggest challenge is to get proper templating and training data inserted into the systems.
AI was never solo-used. It was incorporated as part of the overall content strategy. That means anyway you need a person (writer/expert) that uses AI content suggestions as inspiration and gets the first paragraphs done.
The same competitors use the same AI tools to generate content. Thus, the hardest part is training your own language model by scraping millions of words.
The biggest mistake I see people making when using AI for content generation is failing to provide it with context, relevant data, and original sources. As a result, the output becomes generic and lacks the uniqueness that sets it apart from the rest, just like everyone else using tools like ChatGPT.
Top AI Content Creation Tools
Here are the top 10 AI content generators that SEO professionals use as a teammate before, during, and after the writing process. They can’t do everything for your content ranking, but they can help you be more productive.
The best AI content generators by SEO experts:
- Jasper
- Copy.ai
- Open.ai
- Frase
- Rytr.me
- Writesonic
- Article Forge
- INK
- AI Writer
- Copysmith
- Articoolo
- Hyperwrite
How does AI change SEO
Based on the answers of experts, we can conclude that the topic of the usage AI in SEO is quite controversial, and there are many diverse opinions on the subject, especially for use in the future.
Google will present any entity (person, company, music group, book, product, etc) according to its understanding in the Knowledge Graph. In that space, you can always “beat the machine” since, if you are smart, you will always be the authority with your own entities. Artificial Intelligence/Machine Learning can never fully replace what you know about yourself. Wow.
I don’t know if Google would benefit from determining if the content is written by a human or a robot. The best SEOs are not worried about that. They are focused on whether they are producing the best content to answer the query and build trust and authority. That can be done with or without the help of AI tools.
It’s important to take into account that even if content quality improves and AI-powered tools and features can produce more accurate and comprehensive content with the latest OpenAi’s GPT improvements:
– Can’t be used “as is”, since originality and personalization are needed.
– Real experience just became even more important.
– Semantic and intent optimization is needed.
So, although AI based features will become useful to support and accelerate the copywriting and optimization process, relying completely on it will certainly generate issues sooner than later.
AI tools are not here to replace the writers but to assist them. People will have to ask AI tools for aid sooner or later as it makes your work easy by suggesting multiple options for every content that you generate.
It’s reasonable to think that anything one computer can do automatically, another can detect. And for a human reader (manual check) AI content is easy to detect.
Google is like my mother: she always finds out the truth. So if you are using AI to cheat the system, it is only a matter of time before the truth comes out and you are penalized.
I think this is going to be a cat-and-mouse game. My guess is that Google will likely have the capability to detect AI-generated content and might start penalizing websites at some point, especially without any human edits). AI tools will always play a vital part in the SEO industry because the tools can save a lot of time and resources. Finding the right balance will be the key.
I think over time AI content will be used more and more but you will never fool Google 100% – if you rely on it solely. They have their own AI and Machine Learning models that are more complex…
I think they will be a standard, because it is helping and streamlining the processes a lot, and for Google it will definitely be hard to catch AI content that is edited by humans.
Clever SEOs will use AI as a writing assistant, not as a content creator. Google will learn to detect content that has been written by an AI.
AI Content Generation for SEO infographics
Share this article and infographics with your colleagues and followers on social media to discuss if it’s worth adopting AI in your SEO strategy.
Copy
Conclusion
Admittedly, content creation is a time-consuming piece of the puzzle in the SEO routine. And AI is where the SEO industry is headed. Now, AI content generation tools can be valuable assistants but they hardly substitute human experts.
Ivan Palii
Head of Product at Sitechecker
Ivan joined the team in 2018 and was the first to handle Sitechecker marketing. He had successful results in growing organic search traffic to B2B SaaS websites (Sitechecker, Copywritely).
