A human-centred guide to using ChatGPT for UX research
Pascal Raabe
•
August 23, 2024
•
As UX researchers, we’re constantly on the hunt for faster, smarter ways to uncover rich user insights and inform impactful design decisions. Welcome to the new frontier of UX, where AI can help us wrangle insights from herds of data.
The rise of powerful AI tools like ChatGPT is transforming the research landscape, empowering us to streamline our processes and uncover deeper, more nuanced understandings of user needs.
In this article, we’ll explore practical, step-by-step advice on how to leverage ChatGPT to streamline your UX research—from planning your studies to synthesising findings.
ChatGPT has quickly emerged as a trailblazer in the world of generative AI, capable of processing natural language, generating coherent text, and even providing creative ideas.
As UX researchers, we can harness these capabilities to our advantage in several ways.
Speed and efficiency
ChatGPT can rapidly produce content, from drafting interview questions to summarising qualitative data. This can save us hours of manual effort, allowing us to focus on higher-level analysis and synthesis.
Creative boost
The AI’s language processing abilities can spark new research directions and prompts that we may not have considered on our own, expanding the boundaries of our investigations.
Language analysis
ChatGPT can dive into large volumes of textual data, such as interview transcripts, and uncover trends and insights that may be difficult for humans to spot during manual analysis.
Of course, as with any powerful tool, ChatGPT also comes with certain limitations we must keep in mind.
Contextual understanding
While adept at processing language, the AI may struggle to grasp the full context and nuance that a human researcher would pick up on. Its outputs may require additional refinement and interpretation.
Bias and accuracy
ChatGPT’s responses are based on the data it was trained on, which may contain biases. We must be careful in ensuring these biases don’t seep into our research findings.
Privacy and ethics
Incorporating AI into our research workflows means handling sensitive user data. It’s crucial that we understand the privacy implications and maintain compliance with data protection regulations. More on this, below!
Get 10x more research done with Askable, without the hassle
Let's chatLet’s dive into a specific example of how we can use ChatGPT to enhance our user interviews.
First, armed with our research brief and objectives, we can leverage ChatGPT to help us draft a comprehensive discussion guide.
We might prompt the AI like this:
“Based on this research brief, can you provide a detailed discussion guide covering the key topics and open-ended questions we should explore in our user interviews?”
By having ChatGPT generate the initial guide, we can save significant time and ensure we cover all the critical areas of investigation.
Next, after conducting our user interviews with Askable Sessions, we can grab the interview transcript and upload them to ChatGPT for analysis.
We might prompt the AI as follows:
“I am conducting user interviews with customers of Company X. The objectives of the research are: [insert your research objectives]. Given the research objectives, analyse this interview transcript and provide a concise summary of the key insights, including relevant quotes from the participant.”
The AI will review the text, identify the most salient points according to the research objectives we provided, and deliver a succinct summary—a huge time-saver compared to a manual review.
We can repeat this process for each interview, compiling the individual summaries into a separate document for synthesis in the next step.
Simply copy and paste the output from ChatGPT into a new text document, like this:
“Participant 1: Summary of findings. Participant 2: Summary of findings.”
To identify overarching themes and patterns, we can use this document and prompt ChatGPT as follows:
“I am conducting user interviews with customers of Company X. The objectives of the research are: [insert your research objectives]. Given the research objectives, synthesise the key insights from the interview summaries and identify the 3-5 most impactful themes that emerged. Provide supporting quotes for each theme.”
By leveraging the AI’s natural language processing capabilities, we can quickly uncover the most meaningful user insights, elevating our research findings.
With our insights in hand, it’s time to transform them into compelling, actionable deliverables. ChatGPT can be an invaluable collaborator in this process as well.
First, we can ask the AI to generate an executive summary of our research:
“Can you provide a 250-word executive summary highlighting the key user insights and their implications for product design and development?”
ChatGPT will distil the essence of your findings into a concise, polished overview—perfect for sharing with stakeholders.
We can also leverage the AI’s writing abilities to draft detailed research reports, personas, and other assets.
Simply provide the AI with an outline and key points, then refine the language and structure as needed:
“Using the insights from my research, can you help me write a 500-word user persona for our target customer segment?”
By offloading the time-consuming task of drafting these documents, we free up valuable time to focus on interpreting the data and deriving meaningful recommendations.
Run end-to-end research with Askable and get to insights 10x faster
Contact salesAs powerful as ChatGPT may be, it’s crucial that we maintain a human-centred approach in our UX research.
AI has limitations when it comes to contextual understanding, bias, and accuracy—areas where our expertise and direct user insights remain essential.
As we integrate ChatGPT into our workflows, we must be careful in cross-referencing its outputs against our knowledge and continuously assess for potential biases.
Additionally, we must ensure we uphold strict data privacy and security protocols when handling sensitive user information, for example by redacting the transcripts before we upload them to ChatGPT (e.g. search and replace company and participant names).
Ultimately, ChatGPT should be viewed as a collaborative partner, not a replacement for our skills and judgement as UX researchers.
By striking the right balance between human and artificial intelligence, we can unlock unprecedented efficiency and innovation in our work.
The rise of generative AI tools like ChatGPT are a game-changer for UX research, offering us a wealth of opportunities to streamline processes, boost creativity, and uncover richer user insights.
By thoughtfully incorporating these technologies into our workflows, we can transform the way we plan, conduct, and synthesise our research—all while delivering greater impact for our organisations.
If you’re a UXer who is beginning their AI journey, embrace a mentality of experimentation and continuous learning. Regularly reassess your use of ChatGPT, refine your prompting techniques, and stay attuned to the evolving capabilities of this transformative technology.
After all, one of the most critical skills for a UX researcher is the ability to ask great questions. And this applies just as much when prompting an AI assistant like ChatGPT as it does when connecting with human participants.
