How to use ChatGPT in UX design
ChatGPT can be a useful tool for anyone who produces text-based content as part of their job. It can help with SEO keyword research, SWOT analysis and all kinds of marketing tasks. But what about UX design and research? We’ve been experimenting and here’s what we’ve found.
A bot cannot understand a real user
ChatGPT was designed to be a kind of chatbot. It uses natural language processing to respond to questions and prompts in a way that sounds like a human.
It formulates its responses (to put it very simply) by searching the internet for relevant information. So if you ask ChatGPT how customers feel about a particular brand, or what problems they encounter when using a website you’re redesigning, its answer will be based on data that’s already out there. It will be generic – and possibly inaccurate.
ChatGPT is great at communicating research findings but not so great at research or at genuinely ‘understanding’ it. It can suggest ideas for interview questions, and it can organise and present your findings – writing up a user persona, for example. What it can’t do (not yet, at least) is actually talk to and truly get to know the user.
It can do content
UX people do lots of writing – presentation decks, annotating design libraries, micro-copy, etc. ChatGPT can revise and improve what you write, checking grammar, removing repetitions and so on. You can speed up the process with custom GPTs, which let you save and preload prompts such as “Use UK spelling” and “Revise text” instead of starting from scratch every time.
It's also good for structuring, categorising and organising content. It can help with creating content spec sheets, for example, categorising web pages according to their purpose and description.
And it’s a huge time-saver when it comes to pulling out key ideas from long texts. You can upload interview transcripts and ask, for example, “What challenges did the client mention?” or ask it to bullet-point the key takeaways from a PDF presentation. It’s not always perfect first time – there’s always revising to do. But it’s a good place to start.
Data analysis and coding
Analysing data is often a time-consuming task that involves spreadsheet formulas. ChatGPT can take away a lot of the pain. Just upload Google Analytics data, for example, and query ChatGPT (“What percentage of users churn on the checkout for Canada?”).
And, if you only have basic coding skills, you can use ChatGPT’s guidance to build algorithms using Google’s Apps Script. Again, the results are not perfect, but it is quicker than learning a whole new coding language.
Is ChatGPT after the UX designer’s job?
AI is a helpful tool for UX designers, as long as you recognise its limits. It can present and organise content and help you analyse data, but it cannot do real-life, on-the-ground user research. That’s something that we humans alone can do.
Head here for more on AI in UX research, design and marketing.
Want to stay ahead of the curve with monthly updates on creativity, marketing, and technology trends?
Sign up here to get insights delivered straight to your inbox.