The use cases for ChatGPT in B2B Tech seem to continuously grow, and in digital marketing, the use of ChatGPT in SEO disciplines such as metadata, content creation, and much more has grown tremendously. Our Firebrand Labs team’s mission is to test generative AI tools and determine where they may fit best for our clients. 

We put ChatGPT to the test to understand where the tool fits into your SEO martech stack, identifying the best ChatGPT prompts for keyword research, and developing best practices for using ChatGPT for keyword research.

Getting Started with SEO Keyword Research with ChatGPT

The first thing to note is that the goal of this test is not to use ChatGPT to replace traditional existing SEO keyword research tools, such as SEMRUsh, Ahrefs, and others. That is primarily because ChatGPT does not currently have the ability to provide crucial keyword metrics that SEOs heavily rely on, such as: search volume, keyword difficulty, competitive density and more.

Instead, ChatGPT’s vast training data allows for it to act as a free sandbox for keyword research to be paired with and used with other SEO tools. Just bear in mind that its usage similar to other SEO tools should not be relied on as a single mechanism for approach. 

With that said, SEO keyword research using ChatGPT does provide several capabilities that can help digital marketers expand and improve their keyword research process:

ChatGPT Keyword Research for Topic Ideation

Say you’re getting started on a new blog post – more often than not initial keyword research is the first step to unlocking subtopics to build the outline and headers. However, one downside to this approach is that competitors are using similar SEO tools and consequently targeting similar and often the same keywords. To differentiate your content and expand your keyword profile, ChatGPT can help identify additional subtopics.

Here is an example of an effective ChatGPT keyword research prompt to help generate keyword ideas based on subtopics:

Prompt #1: What are the 5 most popular subtopics related to Account Based Marketing (ABM) in B2B?

This first prompt is to help identify subtopics related to the larger topic, which can in this case be considered a seed keyword for SEO research purposes (account based marketing). ChatGPT’s answer then identified subtopics which we can use to find keyword ideas. 

Let’s take one of the defined account based marketing subtopic examples of “Technology Stack for B2B ABM” to have ChatGPT list out popular keywords in that space.

ChatGPT's response to our first prompt

Prompt #2: List the top 10 most popular keywords for the subtopic listed above “Technology Stack for B2B ABM”

From these keywords results, we can even give further instructions to ChatGPT to obtain long-tail keywords and semantically related keywords which can be really useful for on-page optimization and content creation.

ChatGPT's response to our second prompt

Prompt #3: List the top 10 most popular long-tail keywords for the subtopic listed above “Technology Stack for B2B ABM”

ChatGPT's response to our third prompt.

Prompt #4: List semantically related keywords for the subtopic listed above “Technology Stack for B2B ABM”

Here, the semantically related keywords in the output all contained “B2B ABM” long-tail variations. Say we want broader semantic keywords and concepts, let’s follow up with a follow up prompt to obtain these.

ChatGPT's response to our fourth prompt

Prompt #5: Provide me the semantically related keywords without “B2B ABM” contained

Now there’s a much better result. The outputted semantic keywords can be used in content creation tools and help with further research ideas, such as further queries into entities in SEO  – which ChatGPT is also able to help with! The outlined process can be repeated for any other listed subtopics from the initial output in “Prompt #1” above.

ChatGPT's response to our 5th prompt

ChatGPT Keyword Research Prompts for Maximum Optimization

This launch point for topic based keyword research is only one of many use cases for how to use ChatGPT for SEO keyword research – and we have identified some must-have prompts for various approaches.


The Keyword Brainstorming and Categorization method:

Going back to back our original “Prompt #1: What are the 5 most popular subtopics related to Account Based Marketing (ABM) in B2B?” – we can follow up with a keyword research ChatGPT prompt that will allow us to build out a list of potential “seed keywords” that can be utilized in other keyword research tools to build out lists further and obtain SEO metrics.

In this method using ChatGPT, follow up Prompt #1 above with the following.

Prompt #2: For each of the subtopics, generate a short list of very broad account based marketing keywords.

*ChatGPT generated results for all 10 subtopics – image is cut off here intentionally for sizing.

Now that we have this list, let’s combine ChatGPT’s result with SEMRush to expand on our keyword brainstorming and obtain volume and keyword difficulty. Then, we can go back to ChatGPT to help add some information for these terms.

After pasting them into SEMrush (or your favorite keyword research tool) and obtaining search volume and keyword difficulty, come back to ChatGPT with the following prompt:

ChatGPT's response to our second optimization prompt

Prompt #3: I am pasting keywords with volume and keyword difficulty. Please add a column for: the subtopic from what we did in this chat that would be associated with that keyword, the search intent of that keyword, and a column for additional terms I might want to include in a blog post about that topic: Volume Keyword Difficulty

Marketing automation 14800 83

Predictive analytics 9900 88

Content syndication 2400 56

Ideal customer profile 1600 71

CRM integration 1600 61

ABM strategy 720 57

Attribution modeling 590 55

Content mapping 590 32

Feedback mechanisms 590 57

ROI tracking 480 19

*Note: For the purpose of this blog, we included the top 10 keyword results by search volume for space and sizing.

A table of keywords produced by ChatGPT following our prompt

*ChatGPT’s table included results for all 10 subtopics – image is cut off here intentionally for sizing.

The result here was extremely useful and impressive – and quite possibly the most useful output we have received from any of the tested ChatGPT prompts for keyword research. With this table, we not only have the original subtopics sorted back into the research, but also have additional search intent categories to cross-reference that can help with content creation type. More importantly, the “Additional Terms” provided are accurate for each seed keyword, and provide important related keywords to be included in on-page optimizations for corresponding new content creation.

The Alphabet Soup method: 

Here, we will test using ChatGPT to generate keyword ideas similar to the function of Google autocomplete.

A Google search for account based marketing tactics

Prompt #1: Provide popular keywords that includes the keyword “account based marketing”, and the next letter of the word starts with “t”

Once this result is complete, we can follow up to obtain keyword results containing our focus keyword “account based marketing” as Google autocomplete results for all other letters of the alphabet.

ChatGPT's response to our first 'alphabet soup' prompt

Prompt #2: Move on to giving 5 popular keywords each that include “account based marketing” for every other letter of the alphabet “a to z.”

*ChatGPT generated results for all letters of the alphabet – image is cut off here intentionally for sizing.

ChatGPT's response to our second 'alphabet soup' prompt

The User Persona method: 

Another method of using ChatGPT for keyword research that we found to be quite successful was user-persona based prompts to generate keyword lists. This can be really useful in generating SEO keyword lists for content creation that contains persona-based marketing messaging – for example in a landing page of a gated resource.

Prompt #1: For the topic of “account based marketing” list 10 keywords each for the different types of user personas

The initial result did not include account based marketing specific keywords for each persona, so we followed up with a query to obtain a better result.

Prompt #2: The results under each persona are account based marketing keywords, please adjust.

Much better! The result here provides keywords that are related queries to each of the three accurately defined personas for account based marketing. This can be used in various ways to optimize or generate persona-specific content that speaks to various users and their use case for the topic in consideration.

ChatGPT's response to our user persona prompt

The Search Intent method: 

Understanding the keywords your target persona may be searching is the first step to effective keyword research. The next step is to understand the search intent behind those keywords and which content format may work best. In this example, a “Decision Maker” who is new to account based marketing may search for “what is account based marketing.” On the other hand, if they are further down the funnel and have more intent they may search for “account based marketing agencies.” We can generate a keyword search prompt with ChatGPT to brainstorm core content ideas based on any topic and target user persona. Note that this prompt should be followed up in succession to Prompt #2 above identified in the “User persona method” above for ChatGPT to reference the personas in the previous result.

Prompt: For the topic of “account based marketing” list 10 keywords each for the different types of searcher intent that a “Decision Maker” would be searching for

ChatGPT's response to our search intent prompt

The Pattern method: 

A common SEO strategy is pattern-based keyword research, which leverages variables from a seed keyword to identify important related terms. Below are examples of patterns that we can use ChatGPT to help identify in SEO keyword research:

1. Questions: We can use ChatGPT the same way we would use a traditional keyword research tool to generate various long-tail iterations of the seed keyword.

Prompt: Generate keywords for the topic “account based marketing” that contain any or all of the following “who, what, where, why, how, can, does”

While some of these results are long-tail questions that contain keywords, almost every result can be used in combination with a keyword research tool to generate keyword variations to expand keyword lists. Overall, this was quite useful. 

ChatGPT's response to our first 'pattern method' prompt

2. Comparisons: Similarly, we can use ChatGPT for keyword research to identify comparison patterns and corresponding keywords.

Prompt: Generate 30 keywords for the topic “account based marketing” that contain any or all of “vs, for, best, alternative, review, top”

While all results here are useful, the inclusion of “for” was included here more than other variations. For more comparative keyword results, remove “for” from the list. In addition, the prompt can be adjusted to only include one modifier for more specificity, such as “vs.” Also note that we tested the prompt to generate 30 keywords, however ChatGPT’s token limit will allow for more results – so feel free to expand to up to 100 in the queued prompt.

ChatGPT's response to our second 'pattern method' prompt

3. Branded Identification: Another pattern method that is useful here is to include modifiers in the keyword research prompt that is brand-specific. For certain topics, this can be extremely useful. 

Prompt: Provide the top “vs” keywords for the top account based marketing brands

This prompt can be really useful especially in generating keyword lists for content tailored towards more bottom of funnel user personas that are comparatively looking at [topic] brands and solutions.

ChatGPT's response to our third 'pattern method' prompt

4. Search Intent: Another common method of keyword research is identifying search intent by modifiers that trigger specific variations. An example of this is combining a trigger term in combination with the seed keyword. We can use the following ChatGPT keyword prompt to help us get started. A popular search intent pattern is including “best” to identify when a user is looking to identify top products, types, or services of a specific topic to perform research or help make a business decision. 

Prompt: What are 10 popular keywords related to “account-based marketing” that contain the term “best”?

ChatGPT's response to our fourth 'pattern method' prompt

The ChatGPT result provided search intent variations including “best” that can be used to build out keyword lists for users that are searching within different practices and in various funnel stages of intent in the topic of account based marketing. Like with any of our provided examples of best ChatGPT keyword research prompts, feel free to adjust the focus topic to that in consideration. 

ChatGPT Keyword Prompts for APIs

For more technical SEOs, another category of ChatGPT keyword research prompts can be used to interface with different keyword tool APIs. 

Let’s look at an example to query the SEMRush API.

Prompt: Can you write a script to query the SEMRush API to: take a large list of URLs, get the “top keywords” for each site on the list, sort all of those keywords by traffic they’re driving for the sites?

ChatGPT does note that the user will need an API key from SEMrush to execute this function. With that said, this query can be successfully used in the SEMRush API to obtain top keywords from sites that a user can include in a list. This can be extremely useful in performing competitive keyword research to content gaps and new content opportunities based on top keywords that have driven traffic to competitor webpages.

ChatGPT's response to our first API query

ChatGPT keyword research regular expressions. Now, let’s look at an example of how to use ChatGPT to help with regular expressions (regex) within Google Search Console (GSC) to identify new keyword opportunities. This can also be really useful in the GSC API to help identify new keyword variations.

Prompt: Create a regex in Google Search console that will show me all of the queries in my account that are questions and include “abm” or “account based marketing”

This prompt will help identify search query variations including our focus topic of “account based marketing” or the acronym variation of “abm.” With that data, SEOs can help make more data driven on-page optimizations or help to identify new content opportunities for better content syndication.

ChatGPT's response to our second API query

Key Takeaways for SEO Keyword Research Using ChatGPT

Keyword research with ChatGPT can be extremely useful for SEOs, especially when used in conjunction with other keyword research tools. As previously mentioned, the primary pitfall of using ChatGPT for keyword research is that the generative AI tool cannot provide quantitative SEO metrics. With that said, ChatGPT for keyword research has numerous use cases that are scalable and can help the keyword research process from start to finish – from topic ideation to keyword categorization when used in combination with other tools that provide SEO metrics. In fact, our mentioned use case for using ChatGPT to identify subtopics, followed by using a keyword research tool to obtain SEO metrics and ending with ChatGPT for keyword categorization and additional insights for keyword list building is our favorite.

ChatGPT’s ability to help with initial keyword research and even write scripts for interacting with popular keyword research tools make it a versatile tool that can help SEOs at various points in the process of research and optimizations. For most effective use, think of using ChatGPT as a powerful digital assistant to be added to your SEO martech arsenal. Through practice and trial and error to build muscle memory with the generative AI powerhouse, ChatGPT can be a key resource in optimizing your SEO workflows.

About the Author

Arman Khayyat is a senior digital marketing professional with deep experience in B2B PPC, SEO and analytics. An Account Supervisor at Firebrand, Arman is focused on PPC/SEO strategy, execution, optimization and reporting.