Integrating generative AI into search presents a compelling value proposition.
If SEOs figure out how to adopt this technology effectively, it could revolutionize the industry and fundamentally change how people interact with search.
While there are still some concerns around the accuracy of AI-generated content, as well as omissions leading to some information being taken out of context, SEO expert Eric Enge explained that the technology is still a fantastic tool for creative inspiration and saving time.
Below is a summary how generative AI can be used to drive your technical SEO strategy, as presented by Enge at SMX Advanced.
One of the impressive aspects of ChatGPT is its proficiency in coding, according to Enge. He explained:
- “Chat GPT gives you a very good starting place for your code. Sometimes, it can be a little buggy and needs to be debugged, but it’s a really good place to start – and that’s how I advise you to use it.”
- “If you use it to generate schema markup, remember, it is just a starting place. So make sure you have someone who knows how to read schema markup to validate that it’s right. Using generative AI just speeds up the process.”
Enge shared an example of a scenario where he used generative AI to create Hreflang tags for different webpages:
- “I had four different languages. It gave a bunch of tags and they all look pretty good. This is an example of how [generative AI] can help speed things up for you.”
Generative AI produces accurate results when used for translation purposes. Enge said:
- “I used it to translate keywords from English to French and it produced a really strong set of results.”
- “This sort of thing could really help from the point of view of operations or speeding up the operations of your business because these are fairly common applications.”
4. Implementing .htaccess rules
Enge added that generative AI can also be used to help speed up the process of implementing .htaccess rules:
- “If you have somebody who’s saying that they’re GoogleBot, but they’re not from one of Google’s IP addresses, that’s generally speaking not a good thing. I don’t know explicitly what harm could come from it, but there isn’t any good that will come from it. So you can it to block those.”
- “You can just routinely do a run through to find the the people who are doing that to you and then just block them. This is a pretty good thing to be able to do.”
Enge recommended using generative AI to get creative inspiration for title tags – but stressed that SEOs should also take into account other factors before publishing anything:
- “It came up with a bunch of great titles as prospects to use. It’s great for brainstorming ideas but I would recommend applying search volume analysis too because we don’t really have that data from ChatGPT yet.”
- “But what this technology does let us do is just get some ideas of what we want from the title from the user perspective – and that’s an important perspective to bring.”
When generative AI is used to create meta descriptions, additional prompts can fine-tune the result until you’re happy with it. Enge added:
- “I just kept throwing these little wrinkles at it and it gave decent suggestions for six different meta descriptions.”
- “The best way to create a meta description is to create something that compels a user to want to click through – because we know it has no influence on ranking. So we want to make it enticing to the user so that they have an understanding of what they’ll get if they click on your result. Generative AI can be used to create examples that might be closer to what we might want to use.”
7. Summarizing content
Enge said generative AI is accurate when summarizing content as it carries out this task solely using content you have entered, leaving no room for confusion, date omissions or false information:
- “It’s a great tool for summarizing content because there are no issues with hallucinations.”
Get the daily newsletter search marketers rely on.
8. Keyword grouping
Enge also shared examples of him using ChatGPT prompts to classify and cluster keywords into groups. He added:
- “It actually did quite well – although it wasn’t perfect – with classifying keyword list and groups based on their search intent for the commercial transaction.”
- “However, for clustering the keywords into groups based on relevance, ChatGPT did a really good job. In fact, I would say it was 100% accurate with the answers it gave me.”
9. Content assistance
Generative AI is also a useful tool for generating content ideas. You can ask ChatGPT to create a list of popular questions related to a topic to get a list of potential article ideas:
- “You can use this technique when you’re just beginning to build out a new topic area on your site. You still need to have a subject matter expert go through the details and definitely don’t assume the list it gives you is comprehensive.”
- “However, it can speed up your brainstorming process and I think this is a pretty interesting aspect of this tool.”
Watch: Using ChatGPT to drive technical SEO
Below is the complete video of Enge’s SMX Advanced presentation.