10 Use Cases for Generative AI To Support SEO

Share This Post

Generative AI, like ChatGPT, Bard, Bing Chat, or Google’s SGE, can be used in several ways to support SEO and SEO-related tasks. In this article, we will share some of those ways, including sample prompts. We have used ChatGPT, but most of these will work with other tools as well.

Generative AI to Support SEO

 

1. Determine keyword intent

Provide a list of keywords and use a prompt similar to this:

“Classify the following keyword list in groups based on their search intent, whether commercial, transactional, or informational:”

2. Group keywords by relevance

Provide a list of keywords, preceded by a prompt similar to the following: “Cluster the following keywords into groups based on their semantic relevance.”

3. Generate a list of questions on a topic

Here is a sample prompt for content ideas related to making furniture: “Generate a list of ten popular questions related to building furniture that are relevant for new woodworkers.”

4. Generate article outline

Try this prompt: “Please generate an article outline about generative AI.”

5. Create potential title tags

Try this sample prompt: “Generate 10 unique title tags, of a maximum of 60 characters, for the following text. They should be descriptive and include the term ‘fine wines’ in them. Follow this with the text you want analyzed.

6. Generate meta descriptions

Provide some text you want to analyze, and follow that with a prompt similar to this: “Generate six unique meta descriptions, of a maximum of 150 characters each, for the following text. They should be catchy with a call to action, including the term ‘fine dining’ in them:”

7. Get text descriptions

You can also get pretty solid content created for you if all of the material facts are extracted from a trusted database. For example, if you have a chain of pizza shops, you might want to have a page for each one on your website.

If you have a database of information about each location, including details such as address, opening hours, and some of the food they offer, then this can work to scale.

As part of the process, provide the data for each location and a sample of the desired output. For example, you might provide a prompt such as “Rephrase the following paragraph for three different locations, but avoid repetition while keeping its meaning. Use the following information for each location: ”

8. Generate a list of FAQs for a block of content

Your prompt might be something along these lines: “Generate a list of X frequently asked questions based on the following content:”

9. Find and source statistics

To test this out, we tried the following prompt: “Generate a list of the top five facts and statistics related to the battle of Plassey, including their source.” Below is an extract of the result:

Generative AI for SEO

 

 

 

 

 

 

 

 

 

 

 

10: Summarize content

Sometimes, it’s useful to get shorter summaries of content you’ve written, and generative AI can help here, too. A sample prompt you can use is “Generate a 15- to 20-word summary of the following content.”

Prompt Generation Guidelines

You can use the above examples as starting points in your own efforts, but it’s also useful to understand a general approach to creating successful prompts. Here are a few guidelines:

  1. Be clear and direct about what you want the generative AI tool to do.
  2. Provide an example of the type of output you’re looking for using a very similar topic. For example, if you’re looking to get a brief description of the history of each of the teams in the ICC World Cup, provide an example of the output for one of the teams and ask the tool to use it as a template.
  3. Define your target audience, including whatever criteria make sense for your purpose (age, gender, financial status, beliefs, etc.). The tools will adapt their output accordingly.
  4. Try something and see if it works. If not, modify it and try again. You can re-enter your entire prompt with modifications or chain prompts together, as discussed in the next point.
  5. When you get output that doesn’t meet your needs, tweak your prompt by adding supplemental instructions in a follow-on prompt. For example, if you get a response that is partially correct, provide some corrective input, such as “Thanks for your response, but instead of focusing on the male lead in the play, please focus on the female lead.” Chaining prompts together in this way is often an easier way to refine the output than typing in the whole prompt again, with modifications.
  6. Describe how you want your output formatted. Do you want paragraphs? Bulleted lists? A two- or three-column table? The more specific you are, the better.
  7. Provide context, instructing the tool to “act as if.” For example, you could start your prompt with “ChatGPT, you’re an experienced B2B marketer who deeply understands technical SEO.”
  8. Expanding on point 7, the general idea is to provide the generative AI tool with as much context as you can. This can include who the response is coming from (as per point 7), who the intended audience is (with as many specifics as you can provide), the desired style of writing, whether you want an expansive or narrowly scoped response, and more. These types of prompts are referred to as super prompts.

Conclusion

Generative AI tools can be used to support SEO tasks, as shown in the examples above. There are many more cases where AI can be used effectively. Having a structured approach to generating the prompts will help obtain more relevant results and refine output.

As artificial intelligence continues to be deployed at scale within search and throughout our global data infrastructures, its rapid adoption and significant impact to society are driving urgent questions about the trustworthiness, challenges, and risks of AI and the data upon which it relies.

More To Explore