Turn 100 product reviews into a customer persona: 4 copy-paste prompts
For anyone who sells a product online. Paste your reviews into ChatGPT or Claude, run four prompts in order, and walk away with what buyers love, the words they use, who your buyer is and what to fix.
OutA customer persona
- Goals and values
- Who they are
- Their life and work
- Challenges and pain points
What you walk away with
Four results, one for each prompt, and each one has a job in your store. These are the prompts behind our reel “How to turn 100 reviews into a customer persona”, word for word, from a project in the AI Automation course.
Before you start
Get your reviews in one place
The prompts can only read what you give them, so collect the reviews first.
- Where
- Export them from your store or your review app, or copy them from the product page.
- How many
- The course run used 100. If they are too long for one message, attach them as a file.
- Whose
- Your own product's reviews, or a competitor's. The same prompts read both.
- Then
- Open a new chat in ChatGPT or Claude and run the four prompts below in order, in that one chat. Prompts 2, 3 and 4 build on the answers before them.
Step 01
What buyers like and dislike
Back comes how buyers feel overall, the thing they like most, and the thing they like least.
Real answer from the course run
Use it in your ads. Lead with what buyers like most, and answer what they like least before they ask.
Prompt 1 starts with the role the automation gives the AI. Swap the highlighted words for your own, and paste your reviews where it says [paste your reviews here].
You are a market research analyst. You analyze customer reviews and assess their language and sentiment. You also assist market research by building customer avatars and extracting demographic and psychographic data. Here is a list of competitor reviews for a personal shaver/beard trimmer. Please analyze the reviews and provide a general sentiment analysis. Treat these reviews as a combined whole in your analysis. Also, in addition to the sentiment analysis, please include the most positive trait as well as the most negative trait of the product, based on the reviews. Be as detailed as possible. ### Reviews - [paste your reviews here] ### Constraints: Do not include pretext or context, only return the sentiment analysis and traits.
Step 02
The words they repeat
Back come the keywords, phrases and themes that show up again and again across the reviews.
Real answer from the course run
- Quality
- Battery
- Attachments
- Blades
- Trim/Trimmer
- Durability
- Hair
- Handle
- Charging/Charger
- Packaging
- Product quality and durability issues
- Battery and charging concerns
- Effectiveness and versatility of usage
- Design flaws
- Quality control and customer experience
- Value for money
Use them on your product page. Describe the product in the words buyers already use for it.
Now provide a list of repeated keywords, repeated phrases, and common themes among the reviews. Be as detailed as possible.
###
Constraints: Do not include pretext or context, only return the keywords, phrases and themes.Step 03
Who your buyer is
Back comes a customer avatar: their goals and values, who they are, and the problems they have.
The shape of the answer
OutYour customer persona
- Goals and valuesWhat they want to achieve, and what that says about their values
- Who they areAge, gender, marital status, children
- Their life and workLocation, occupation, income, education
- Challenges and pain pointsWhat gets in their way, in detail
Use it for who you target. Aim your ads, your content and your offers at this one person.
Now, create a customer avatar based on the content of these reviews. Here is the information we hope to glean from the reviews to create the avatar -
1 - Identify goals and values. What are the goals that my business can help them achieve? What are their commitments and how does that infer their values?
2 - Demographic information. Age, gender, marital status, children, location, occupation, job, income, education...if possible.
3 - List out their challenges and pain points. Be as detailed as possible.
###
Constraints: Do not include pretext or context, only return the customer avatar.Step 04
Why they buy, and what to fix
Back come the reasons people bought, and suggestions to improve the product based on what they said.
The shape of the answer
- Why they boughtThe purchase motivations in the reviews
- What to fixSuggestions to improve the product, based on the feedback
Send it to your supplier or your next version. Fix what buyers complain about, and keep what made them buy.
Finally, list the purchase motivations inherent in these reviews and provide suggestions on how this product could be improved based on the feedback. Be as detailed as possible.
###
Constraints: Do not include pretext or context, only return the purchase motivations and suggestions.Let it run on its own
The same four prompts, as one automation
- AI reads
- Insights
- Emailed
In your emailThe report
- Review Sentiment Analysis
- Repeated Keywords, Phrases, and Common Themes
- Customer Persona
- Product Improvement Suggestions
In the AI Automation course, you build this as an automation. You upload the file of reviews, the four AI steps run one after another, and the report lands in your email with a section for each prompt.
It is Review Sentiment Analysis, one of the three projects you can choose at the end of Module III (project 3.5.2).
The prompts and the run are from Project 3.5.2 of the AI Automation course, taught by Anthony Lee.