ChatGPT keyword research: actual prompts
Chat GPT keyword research tools, and a framework to incorporate the tool into your SEO process.
Chat GPT is a tool that allows you to create content for SEO.
- It is a great time-saver and can be used for certain tasks.
- It’s worse than what you do for other tasks.
- It is important to understand the system and be able to modify your prompts, sometimes several times, and to QA the output.
ChatGPT keyword research is easy with all of these facts. Let’s take a look at the steps to help you understand them.
- Chat GPT excels at keyword research (with certain prompts).
- Chat GPT isn’t good at keyword research functions
- Chat GPT – How to QA keyword research output
This article will provide you with ChatGPT keyword research tools and a framework to incorporate the tool into your SEO processes.
ChatGPT: How to Use it for Keyword Research
Chat GPT doesn’t have the ability to access search volume or other metrics like keyword tools such as Ahrefs, Google Keyword Planner, and Semrush.
Chat GPT is a great tool for keyword research.
Topic ideation and keyword brainstorming
Traditional keyword research tools have a downside: your competitors, and often more authoritative sites, are using the same tools and targeting similar keywords.
The right Chat GPT prompts will help you discover largely untapped keywords as you create a list.
Topic for dummies
Chat GPT could provide some ideas and prompts for me to create a website on pickleball.
This is not a list of keywords. However, it can serve as a site structure and give me ideas for building keyword clusters.
Agendas for the Topic’ conference
ChatGPT can also be used to ask for conference topics that target a particular persona within my niche. This provides interesting ideas.
Competitive research done quickly and easily
ChatGPT is not a keyword research tool and doesn’t provide the most current data. It can give me a quick overview of the niche’s competition and help me get started with high-level concepts.
You’ll also find more ideas on topics and sub-niches.
Research on influencers and social media
You can keep up with the latest news in this niche by checking out social media accounts on platforms such as Twitter.
Keep in mind that this data comes from the GPT-3 training sets, which aren’t current. Twitter also suspended access on December 4, 2022
Reddit is also available:
You can make this even more helpful if you dig a little deeper.
You can get great output by using specific angles to create articles or FAQs.
You can also see a quick overview of thought leaders and influencers:
This allows me to explore more topics and learn about the topics they tweet.
Next, I can begin to put this all together by combining prompts and asking ChatGPT for a summary of its research.
I will be completely honest and say that I do a lot of ChatGPT prompts. Some of the output is horrendous.
It would be okay if a VA was hired to research a niche and then asked them to provide some ideas for high-level topics and categories for a new website.
Create actionable keywords
Now we have data on high-level categories as well as “sub-niche” information. How can we turn this into a list of targeted keywords?
ChatGPT cannot give you an estimate of search volume or keyword difficulty. They aren’t perfect metrics but you don’t want to waste your time or resources if they’re not accurate.
ChatGPT can be used to create a list “seed keywords” that we can run against our favourite keyword tool.
You can refine your results based on what you receive back.
Next, I used a prompt (“Perfect, take those keywords and transform them into a list that I can easily copy/paste”) to clean them up. I then dropped them all into Ahrefs and looked at matching terms. I then filtered for keywords with a keyword difficulty below 5.
Now that I know the search volume and difficulty of certain terms, I can return to Chat GPT to add more information. It is quite impressive.
You can also find interesting keyword data from my new list by asking “Can you tell us how many posts are within each sub-topic?”
You may find it useful to balance your content output across business lines, categories and support for different tools and products. This allows you to quickly sum up the information.
Keyword list expansion, cleanup and generation
Keyword research is critical in finding the right keywords and adding modifiers to them to increase your list.
One common use is to create a list of geo-modifiers, and then marry them with a target term.
You can also obtain a list with size modifiers that you can apply to different terms.
There are some very interesting ideas in the output:
The mash-up process is easier if you have seed lists of modifiers.
Keyword research at the page level
Many tools can create content briefs or grades for you content based on search engine rankings. These tools can be quite expensive and I use several of them in my daily work.
ChatGPT can be used to compare the article to other articles ranking for certain terms. It is not an SEO tool native to the site, so data may differ from those used by other tools.
ChatGPT does not crawl URLs yet. To have Chat GPT analyse your articles, you will need to copy and paste the article ranking and article title into the chat. This can be frustrating as the limit of characters is 2,000 words.
Here’s the analysis
You can also get more detailed advice by following up with a prompt
ChatGPT can remember past prompts in a chat.
I could then run through several articles ranked on the first pages for this search term. Chat GPT would pull out common terms and themes from each article’s headers, and then surface most popular phrases in these articles.
Manually doing this is a bit cumbersome as some pages or articles will have more than 2,000 characters. Chat GPT will attempt to answer you before all content has been posted.
It’s not answering my question, as you can see from the above. This is where the tools that were specifically made for it are much more user-friendly.
Open AI API allows for a smoother, more scalable experience with prompt-heavy tasks. ChatGPT will generate the code.
ChatGPT prompts keyword research APIs
Remember to be cautious when working with programmatic APIs that are expensive or require credits. You should inspect everything that the tool outputs.
ChatGPT also offers another keyword research function: It allows SEOs to interface with different keyword tool APIs.
I was able to grab a large list of sites with low domain authority that rank for pickleball terms, and run this script on them.
You can ask the tool for step-by-step instructions if you are not technical.
Although not relevant to my pickleball website, ChatGPT can help me with keyword research using the Google Search Console API. I can find search queries that might require their article or page by asking ChatGPT:
ChatGPT is not the only way to obtain keyword data via GSC API.
ChatGPT keyword research regularly used expressions
Regular expressions (regex), similar to the GSC API can be a great tool for finding keyword targeting opportunities.
Make your own tools
ChatGPT will show you how to create specific tools for keyword research. Always ensure quality assurance and be cautious!
ChatGPT Keyword Research QA Tips
These examples show how useful Chat GPT is for keyword research. It is available for free up to $20 per month. This makes it worth the effort.
My experience shows that the platform is not very good in key areas such as keyword research.
- Chat GPT is not search-specific. It’s a native keyword research tool. It doesn’t have access to keyword data directly and cannot perform key functions in keyword research.
- Confident, but incorrect: Sometimes the platform provides false answers. It presents data as facts but it is completely wrong.
- Broken code: ChatGPT code does not always work. It is possible for it to interface with an API, or to build a tool that automates certain aspects of your process.
These are some suggestions that might help you:
- When working with ChatGPT code and instructions, ensure you use dev environments.
- To ensure you aren’t wasting your time on topics that don’t make sense, check any keyword suggestions against competition and search data.
- Ask the tool to analyse something. Take the output with a grain. ChatGPT can be wrong, cannot crawl URLs, and has outdated data.
Broader Chat GPT keyword research strategies
Chat GPT is a great tool for keyword research, but there are caveats. These are the questions you should ask to help you decide how to best use ChatGPT.
- What can you do with the Chat GPT API to your advantage?
- You can use the API to perform various functions if the chat character limit is too restrictive or you need to search for a topic or keyword that has a greater scope.
- Do you have other apps or APIs that you would like to use, but don’t have the time or energy? You could try a mashup of multiple APIs (marrying data e.g. from Ahrefs, GSC or Semrush).
- What are the most time-consuming and tedious keyword research tasks you have to do? These are the tasks Chat GPT can help you with.
Chat GPT is a virtual assistant. You are ultimately responsible for any code or data it generates. You should think about the tasks that you have to do and the possible capabilities of the device. Then, try them out when they make sense.