Know who you/the client is talking to:
Caveat: This is 1000% not as powerful as conducting research directly with your target audience via a client discovery session, user focus group, or a “jobs to be done” formal research session. Ideally, you confirm what you get from AI with human conversations with your target audience.
But if you’re lost as to who you are talking to, this is a better starting point than not thinking about your target audience. I’ve enhanced my basic prompt here with tips from Andy Crestodina to make it a chained set of prompts:
Build me a persona of a [job title] who works for a [company type] company. List the roles, goals, challenges, pain points, and decision criteria for selecting a [your company type] company.
What keeps that persona up at night?
What words does that persona use to describe his/her [service] challenges and solutions?
If you were [describe your persona], what online resources would [persona] read to help you be more effective at your job?
Topical research — What topics should you cover? What are your topical gaps?
Here are the titles of our published blog posts related to X content. What content is missing?
Based on the episodes in this feed, [feed URL], what topics related to “[keyword]” have not been covered? The target audience is X, and the content is focused on X or helping X persona with x.
Know the main topics and subtopics that must be covered in your copy
Create a table with the semantically related subtopics of X topic organized by the stages of the searcher’s learning about the topic. The output should have one column for the learning stage and another with the keywords used in a search.
I saw this one shared by Aleyda Solis’s SMX Advanced presentation and have adapted it to my workflow.
Export a raw list of keywords and their search volumes from Moz Pro’s Keyword Research tool. Upload this list to ChatGPT and ask the AI tool to cluster the keywords into semantically associated groups (while maintaining the associated search volumes). Export this to a .csv file.
Create a hub and spoke model from a list of these keywords, consolidating those that are too similar and would be targeted by the same content to avoid duplicates.
Take their search intent into consideration to group them.
These will be used in the X section on our website.
The hub and spoke model should be in English and list each keyword’s search intent.
The target audience is [x], and this should attract and engage [name of the target audience here] for a [type of business/service].
List them in a table along with their specified average search volume in additional columns listed below.
Additionally, suggest titles and meta descriptions following SEO best practices for creating content to target them: [list keywords here].
Know your unique brand value and how to surface it in content
This needs to be executed with an AI chat with access to the internet (ChatGPT4 or Bing Chat) or ChatGPT with the BrowserOps plugin:
Based on these competitors [list competitors names & URLs], how is [your brand + URL] different?