This guide will walk you through a step-by-step process for transforming raw survey data into actionable insights using ChatGPT.
So let’s get started…
Setting Up ChatGPT
If you do not have an account yet, click here.
Privacy Best Practices
- ➡️ Head to the Settings menu and look for the section labelled Data Controls.
- ➡️ Toggle the setting for ‘Improve the model for everyone’ to disable it.
- ➡️ For a recap on how to use ChatGPT for Market Research safely, click here.
Resources for Practice
We’ve included a downloadable synthetic dataset, a prompt document, and a questionnaire to help you follow along and practice each technique.
Download the Files 👇
The files are securely hosted on Internext.com and no account is needed. If you prefer, I can email them to you. Simply get in touch!
- 📊 Dataset
- ✏️ Prompts
- 📄 Questionnaire
Getting Started
To begin:
- 📥 Download the dataset.
- 🗂️ Open a new ChatGPT conversation.
- ✍️ Attach (upload) the dataset and provide a brief description, such as:
“This is a survey dataset containing questions on people’s perceptions of punk music.” - 🔄 Copy and paste the starter and follow-up prompts from this guide to learn the processes and techniques.
💡 TIP: Copy all text located between the two asterisks (*) for prompts.
By the end of this guide, you’ll learn how to:
- Add enriched columns to your dataset.
- Make your data more actionable, visualisable, and insightful. 🚀
🔗 Step 1: Connecting the Dots Between Questions
Scenario
You want to explore relationships between satisfaction, sustainability perceptions, and shopping frequency.
Starter Prompt
*Combine the following:
Satisfaction ratings (Q1),
Sustainability associations (Q2),
Shopping frequency (Q9).What patterns or correlations emerge? Highlight differences between frequent (4–5) and infrequent shoppers (1–3).*
ChatGPT Output
The output identifies patterns or correlations, highlighting differences between frequent and infrequent shoppers.
Follow-Up Prompt
*Add a column in the dataset called Sustainability_Perception_Group to categorise responses as:
Positive (e.g., ‘eco-friendly materials’).
Neutral/Negative (e.g., ‘too expensive’).*
🔮 Step 2: Predicting Future Loyalty
Scenario
Predict customer loyalty based on satisfaction and recommendation likelihood.
Starter Prompt
*Using the following:
Satisfaction ratings (Q1).
Recommendation likelihood (Q4).
Shopping frequency (Q9).
Predict how customer loyalty might change over the next six months. Consider external influences like affordability perceptions (Q3).*
ChatGPT Output
The output analyses relationships between variables, predicting loyalty trends and potential changes over six months.
Follow-Up Prompt
*”Add a column called Predicted_Loyalty with categories:
High for frequent shoppers with high satisfaction.
Medium for infrequent shoppers who recommend.
Low for hesitant recommenders with low satisfaction.”*
🌍 Step 3: Uncovering Cultural Attitudes
Scenario
Explore cultural or generational attitudes toward sustainability.
Starter Prompt
*Analyse responses to ‘What comes to mind when you hear sustainability?’ (Q2). Categorise responses into themes like:
Affordability Concerns
Eco-Friendly Enthusiasm
Skepticism.*
ChatGPT Output
The output groups responses into themes, uncovering patterns and organising them into categories.
Follow-Up Prompt
*Add a column called Sustainability_Themes with categories suggested.*
🧠 Step 4: Identifying Hidden Biases in Data
Scenario
Evaluate if survey questions skew responses.
Starter Prompt
*Evaluate these survey questions for bias:
‘Why do you shop with us?’ (Q5).
‘Rate affordability vs. quality’ (Q3).Do these questions assume certain preferences? Suggest neutral alternatives.*
ChatGPT Output
The output identifies potential bias, suggesting neutral alternatives to avoid skewed responses.
Follow-Up Prompt
*Annotate biased questions in the dataset metadata and suggest neutral revisions.*
🎭 Step 5: Creating Personas
Scenario
Create personas based on survey results.
Starter Prompt
*Using satisfaction ratings (Q1), shopping motivations (Q5), and improvement suggestions (Q6), create three personas with:
Motivations
Barriers
Ideal messaging recommendations.*
ChatGPT Output
The output synthesises responses into three distinct personas, offering tailored messaging recommendations.
Follow-Up Prompt
*Using the dataset, assign each respondent a persona in a new column called Assigned_Persona.*
⚖️ Step 6: Exploring Contradictions
Scenario
Explore conflicts between quality and affordability ratings.
Starter Prompt
*Analyse contradictions in:
Q3: Affordability vs. Quality.
Q7: Purchase decision factors.
Why might respondents prioritise quality but select affordability as a key driver?*
ChatGPT Output
The output analyses contradictions, exploring reasons behind trade-offs and balancing priorities.
Follow-Up Prompt
*Add a column called Affordability_Contradiction_Reason summarising these insights.*
❤️ Step 7: Tracking Emotional Journeys
Scenario
Track emotional responses to eco-friendly initiatives.
Starter Prompt
*Classify emotional responses (Q8: Excited, Skeptical, etc.) into Positive, Neutral, or Negative sentiment.
Assign scores:
Positive = 5
Neutral = 3
Negative = 1What patterns emerge?*
ChatGPT Output
The output categorises responses into sentiment groups and scores, identifying trends across respondent groups.
Follow-Up Prompt
*Add two columns:
Sentiment_Category (Positive, Neutral, Negative).
Sentiment_Score (Numeric score).*
💡 Final Thought
This guide demonstrates how ChatGPT can turn survey data into actionable insights. Whether predicting loyalty, creating personas, or uncovering contradictions, these techniques will help you unlock the hidden power of your data.
Use this as a starting point to refine your approach, explore advanced techniques, and discover even deeper insights.
This training guide is yours to use however you like. Share it, apply it in your business, or adapt it to fit your brand. That is what DIY culture is all about! If you find it valuable, consider supporting my work with a coffee or a beer through my Buy Me a Coffee link below. Cheers! ☕🍻
