.svg)
UX research and generative AI tools, like ChatGPT, are a powerful pairing for insight discovery. To work your research magic, we're sharing 5 advanced prompts that conjure clear, impactful insights in no time.
As UX professionals, we’re accustomed to asking the right questions and employing robust analytical processes to generate insights.
Now, it’s time to apply those skills to a new frontier: engineering AI prompts to get insights from user interviews with ease and accuracy.
This article will guide you through a series of advanced ChatGPT prompts tailored for UX research, demonstrating how AI can be a powerful assistant in your quest for impactful insights.
At its core, prompt engineering is the art and science of crafting effective instructions for AI models like ChatGPT to produce desired outputs.
It’s about knowing how to ‘speak’ to AI in a way that elicits the most useful and relevant responses for your specific needs.
As AI continues to evolve, mastering prompt engineering will become an increasingly valuable skill for UX researchers.
Efficiency
Well-crafted prompts can significantly speed up various research tasks, from generating interview questions to analysing transcripts.
Consistency
Standardised prompts ensure a consistent approach across different research projects or team members.
Depth of insights
Advanced prompts can help uncover patterns or connections in data that might not be immediately apparent to human researchers.
Scalability
With effective prompts, you can process and analyse larger volumes of data more quickly.
Innovation
Creative prompt engineering can lead to novel approaches in research methodology and insight generation.
Customisation
Tailored prompts allow you to adapt AI capabilities to the specific needs and nuances of your UX research projects.
Complementing human skills
By offloading certain tasks to AI, researchers can focus more on interpretation, strategy, and applying insights to design.
Let’s explore some useful frameworks and prompts to help you engineer highly efficient AI prompts for your UX research.
Before diving into the frameworks, meet Ask AI—a new feature from Askable that elevates your AI-assisted research.
With Ask AI, you can instantly get answers from your research transcripts, complete with evidence-based citations. Simply type your question in the session playback screen, and let the AI provide quick, accurate insights, backed by specific quotes.
This feature saves you time on analysis, allowing you to focus on what matters most—uncovering actionable insights and making informed decisions faster.
As UX researchers, interviewing people is our bread and butter. We understand the power of asking the right questions to uncover valuable insights.
Interestingly, we can apply this same principle when working with AI, using it to help us clarify and refine our research objectives.
Here’s how you can use ChatGPT to interview you about your research project and synthesise your research objectives.
Help me identify the research objectives for a UX research project. I want you to interview me by asking 10 highly relevant questions, one at a time, which will help you build a thorough understanding of the project context. Once I have answered your questions, I want you to generate a concise summary of the research objectives.
By engaging in a structured dialogue with the AI, you’re able to explore your project from multiple angles, often uncovering aspects you might have initially overlooked.
This process naturally leads to greater clarity in your thinking, as you're prompted to articulate your project details in response to specific, targeted questions.
The output can then be used to provide sufficient context in your analysis prompts once you have conducted user interviews.
Starting with this AI-assisted objective-setting process helps ensure that your subsequent research activities are well-aligned with your project goals.
The RTF (Role, Task, Format) framework offers a simple yet powerful structure for crafting AI prompts that yield precise, tailored results for your UX research needs.
This approach helps you clearly define the context, specific action, and desired output format, ensuring that the AI’s response aligns closely with your research objectives.
Act like a [insert the role you want AI to take]. Give me a [insert task] in [insert format] format.
This framework is particularly effective for generating specific UX research tools or outputs.
By clearly defining the role you want the AI to assume, you tap into its ability to mimic expert perspectives.
The task specification ensures you get exactly what you need, while the format instruction guarantees the output is in a readily usable form.
Here’s an example of how you might use this framework in your UX research process:
Act like a senior UX researcher.
Give me a set of 10 open-ended interview questions for a study on mobile banking user experience in bullet point format.
This prompt would yield a structured list of thoughtful, probing questions that you could use as a starting point for your user interviews.
The beauty of this approach lies in its flexibility—you can easily adjust any part of the RTF framework to suit different stages of your research process, from initial planning to data analysis and reporting.
Get 10x more research done with Askable, without the hassle
Let's chatThe RISEN (Role, Instructions, Steps, End goal, Narrowing) framework provides a structured approach to dissect complex or constrained UX tasks into actionable components.
This method is particularly useful for more intricate research activities, such as insights synthesis or detailed analysis of user data.
Role: [Insert the role you want AI to take]
Main task: [Insert the task you want AI to complete]
Steps to complete task: [Insert numbered list of steps to follow]
Goal: [Insert goal of the output]
Constraints: [Enter constraints]
This framework allows you to guide the AI through a detailed process and ensures that the output aligns precisely with your research needs.
By breaking down the task into specific steps and clearly stating the end goal and constraints, you can obtain highly targeted and relevant results.
Here’s an example of how you might use the RISEN framework in your UX research:
Role: Expert UX research analyst
Main task: Analyse the transcript of a user interview about a new fitness tracking app
Steps to complete task:
1. Identify key themes and patterns in user feedback
2. Extract specific pain points and areas of satisfaction
3. Note any feature requests or improvement suggestions
4. Highlight quotes that best illustrate important insights
5. Summarise findings in a concise report
Goal: Produce actionable insights to guide the next iteration of the app design
Constraints: Focus only on aspects directly related to the app’s user experience—disregard any off-topic discussions.
The BEAM (Behaviour, Emotion, Attitude, Motivation) framework is an excellent tool for gaining a comprehensive understanding of users.
This approach is particularly valuable for discovery phases and can significantly contribute to the development of rich, nuanced user personas.
Analyse the attached user interview transcript according to these four aspects:
Behaviour: What actions or behaviours are observed in the user?
Emotion: What feelings or emotions are associated with these behaviours?
Attitude: What attitudes or beliefs underpin these behaviours and emotions?
Motivation: What drives or motivates the user’s behaviours and attitudes?
Provide a detailed analysis with specific examples from the transcript for each element of the framework.
This framework encourages a holistic view of user experiences, going beyond surface-level observations to uncover the underlying factors that influence user behaviour.
By prompting the AI to consider each of these aspects, you can obtain a rich, multi-dimensional understanding of your users.
The BEAM framework is especially useful when used in a two-step process—first, analysing individual transcripts, and then synthesising the compiled information from multiple transcripts.
The AI can quickly process this information and extract insights across all four dimensions, providing you with a comprehensive overview that can form the basis of user segmentation or personas.
Run end-to-end research with Askable and get to insights 10x faster
Contact salesThe TNOD (Tasks, Needs, Obstacles, Desires) framework can help us systematically break down user experiences into key components.
It provides a clear picture of what users are trying to accomplish, why, what’s standing in their way, and what improvements they’re hoping for.
Analyse the following user interview transcript and identify:
1. Tasks: What specific actions is the user trying to accomplish?
2. Needs: What are the underlying needs associated with these tasks?
3. Obstacles: What pain points or challenges does the user face in completing these tasks?
4. Desires: What solutions or improvements does the user wish for?
Please provide a structured summary with examples from the transcript for each category.
This framework is particularly valuable in evaluative research, when you’re dealing with complex user journeys or trying to identify opportunities for product improvements.
By separating user experiences into these four distinct categories, you can gain a more nuanced understanding of user behaviour and motivations.
To take your research efficiency to the next level, consider creating custom GPTs. These are personalised versions of ChatGPT that you can tailor to your specific UX research needs.
Log in to ChatGPT. You’ll need a Plus or Enterprise Plan to create your custom GPT.
Click ‘Explore GPTs’ in the side panel.
Click ‘Create’ and enter your instructions in the message bar.
Chat with the GPT Builder to refine your custom assistant.
Use the ‘Configure’ option to add advanced customisations.
Upload your research brief or relevant knowledge files (see note below).
Choose whether to share your custom GPT, or keep it private.
Click ‘Create’ and then ‘Update’ to finalise your GPT.
Note on knowledge sharing: While OpenAI offers privacy controls, including the option to opt out of model training, it’s crucial to exercise caution when using AI tools for research purposes.
Even with these safeguards, there’s always a potential risk of data breaches or unauthorised access to your information.
To protect sensitive data and maintain participant confidentiality, it’s advisable to redact any identifying information from transcripts and documents before uploading them.
This includes removing or replacing company names, participant names, and any other potentially identifiable details.
You are a helpful UX research assistant, capable of analysing user interview transcripts diligently, objectively, and with rigour.
When I say ‘Start’, analyse the attached user interview transcript according to the research brief.
Structure your analysis for further synthesis, focusing on the research objectives outlined in the brief.
Always provide 1-3 quotes directly from the transcript to support your findings.
Of course, your instructions could include any of the frameworks introduced above.
Once you’ve conducted interviews using Askable Sessions and downloaded the transcripts from the platform, you can use your custom GPT to analyse them quickly and consistently.
All you need to do is select your custom GPT, attach your user interview transcript and type ‘Start’.
Your custom GPT will perform an analysis of the transcript according to your prior instructions, and the information you uploaded.
From here, you can continue to explore deeper and have a conversation with your new research assistant about the analysis.
For example, you could ask your custom GPT the following questions to uncover further important insights from your interview:
Remember, you don’t necessarily need to download the transcripts and run them through a custom GPT—you also have the option to simply use Ask AI to get instant, evidence-based analysis within the Askable platform.
AI is a powerful tool that’s well suited to enhance human expertise in UX research.
These advanced AI prompts can enrich your research processes—from planning and conducting interviews, to analysing data and generating insights.
Use these prompts as a starting point, and iterate and refine them based on your specific research needs and contexts.
By leveraging custom GPTs and applying your UX research skills to the art of prompt engineering, you can streamline your workflow, uncover deeper insights, and focus more on applying those insights to create better user experiences.
And with Ask AI, you have a solution that simplifies the analysis process even further, delivering instant, actionable insights directly from your sessions.
