How AI Changed the UX Researcher's Role
โAI in UX research has changed less about the day-to-day tools and more about how your colleagues see your role. Not in the way most people mean when they talk about AI transcribing interviews, but structurally: where researchers once designed the study, senior leaders now use AI assistants to translate their strategic goals straight into named methods, conjoint analysis, MaxDiff, price sensitivity testing, and hand those to the researcher to run.โ
The researcher has shifted from the London Black Cab driver, trusted to plan the route from deep expertise, to something closer to an Uber driver following an AI-generated route as fast as possible. The trade-off cuts both ways: less control over which methods you use, but far more strategic, higher-stakes work. Staying indispensable now means upskilling into the advanced quantitative and mixed methods that leaders are suddenly asking for.
You're not a UX Researcher anymore, at least not in the way you were. LLMs have changed how research projects end up on your plate.
I'm not talking about using AI to summarise user interviews or pull the perfect quote from your research repository. That's small, everyday efficiency stuff. I'm talking about how your colleagues read your role as a UX Researcher, and how they'll work with you from here, which has changed for good since LLMs took off.
Understanding these new rules and expectations matters if you want to thrive as a researcher. What follows is a standing take on AI in UX research, not a reaction to any one product, because the shift is structural and here to stay.
Before AI: the researcher in the driver's seat
Before generative AI arrived, UX Researchers were in the driver's seat. Senior leaders brought them their tough questions and strategic objectives to translate into a solid research plan. They set the destination. You figured out how to get there.
Not anymore. I first noticed the shift during onboarding calls with UX Researchers joining OpinionX:
โI never thought I would be asked to ask end users about pricing. Iโm a UX Researcher and Iโm finding I do โcrossoverโ market research more often than I expected.โ
โAnd now theyโre telling me to run pricing and conjoint surveys, which is the marketing teamโs job, not mine.โ
โI ran [a conjoint survey] through a research agency once before, but now they expect me to design and run and analyse all this work on my own. It feels like Iโm way out of my depth.โ
These leaders hadn't spontaneously enrolled in User Research 101 or binge-watched YouTube explainers after putting their kids to bed. Yet somehow they were confidently requesting advanced research methods by name, methods they almost certainly had never heard of before.
What had changed? Why were UX Researchers suddenly landing advanced research methods for high-stakes quantitative projects, with little notice or support?
The culprit was AI. Before AI assistants, UX Researchers were like London's Black Cab drivers: trusted to use their deep knowledge of every street in the city to find you the best route to your destination. Today, the role is more like an Uber driver, where the most efficient route is already mapped and your job is to follow it, as fast and smoothly as you can.
Your boss now uses an AI assistant to turn their strategic objectives into specific research methods, tools and techniques, bypassing you in the process. But this shift didn't start with AI. AI only accelerated it. Let's look at what led here and what it means for UX Researchers going forward.
How Product-Led Growth made mixed-methods research inevitable
Product-Led Growth (PLG) began reshaping software strategy in the late 2010s. It introduced three shifts:
Free tiers and free trials became baseline expectations.
Purchasing became self-service for entry-level price points.
End users started buying tools to solve their own problems, instead of relying on top-down procurement.
These changes pushed user experience to centre stage. Before PLG, user experience was mostly a way to reduce churn after purchase. Today, UX itself is a major driver of revenue growth, especially in freemium and free-trial models.
โIf the buyer is now the end user, then user experience is your new salesperson.โ
In the early 2010s, UX Research was deeply qualitative, with a toolkit centred on usability studies, user interviews and thematic analysis. Most UX Researchers came from cognitive science backgrounds, with PhDs in psychology and HCI being particularly common.
Once UX got its seat at the leadership table, researchers hit a problem: the leadership team's common language was quantitative. UX Researchers needed hard data, not anecdotes, to help shape big decisions. To stay relevant, they had to adapt. Mixed methods research became the bridge: the qualitative side gave depth and nuance, the quantitative side gave scale and impact.
This shift also made it necessary to bring research in-house. UX had become a continuous driver of growth, not just something you tested now and then. Back then, there were three main types of in-house researcher: the UX Researcher, the Market Researcher, and the Consumer Insights Manager. Each had their own methods, skillsets and focus areas, often with little overlap.
The traditional division of labour across these roles had made sense in the agency world, where large teams allowed for deep specialisation. But in-house research teams were rarely that big. Sometimes a single researcher was expected to support an entire product org, even in companies making tens of millions in revenue. In that context, methodological flexibility wasn't optional. It was a necessity.
As more UX Researchers added quantitative methods to their toolkit, a new kind of researcher emerged, one who could run user interviews and a conjoint analysis survey in the same week. That hybrid earned a new term: Mixed Methods Researcher.
I first wrote about this shift in 2020, predicting that the 2020s would be the decade of Mixed Methods Researchers. At the time, only a handful of people on LinkedIn had "Mixed Methods Researcher" in their title. That number has since grown into the thousands and continues to rise.
The shift from "UX Researcher" to "Mixed Methods Researcher" was, for many, a clever act of professional self-preservation. It became a real factor during the 2022 tech layoffs, where those who had already upskilled in quant methods were more resilient to job losses. They didn't just do more, they could speak the quantitative language of their leadership team.
As mixed methods became the new gold standard in tech, the differences between UX Researcher, Market Researcher and Consumer Insights Manager began to blur. The new archetype was the generalist in-house researcher: one person capable of completing the work that had required three job titles.
Although Product-Led Growth expanded the scope of what UX Researchers worked on, the process within their role hadn't changed much. Colleagues still brought their big strategic questions to the research team, expecting expert guidance on designing the right study. Researchers were still the trusted Black Cabbie, until AI arrived and everything changed.
After AI: the researcher as Uber driver
Once LLMs took off, a subtle but significant shift appeared in how people worked with in-house UX Researchers. I kept hearing the same story from researchers asked to run projects more advanced, more quantitative, and further outside their comfort zone than anything they'd done before.
Tasks that once meant running a simple survey were now conjoint analysis projects. Turning survey results into charts had become segmentation or clustering analysis. Preference testing on design mockups turned into price sensitivity testing. The expectations and complexity kept growing.
UX Researchers weren't the ones choosing these methods. Their role in research design, once a core part of their work, was increasingly bypassed. Instead of the London Black Cabbie, trusted to use deep expertise to navigate any research challenge, the modern UX Researcher is treated more like an Uber driver, expected to follow the AI-generated route as fast as is legally compliant.
AI didn't just change the tools and methods researchers are expected to use (PLG had kickstarted that years before). AI changed how the researcher's role is perceived within the company.
| Before AI (the Black Cab) | After AI (the Uber driver) | |
|---|---|---|
| Who designs the research | The researcher, from expertise | AI suggests the method; the researcher executes |
| What leaders bring | A strategic question to translate | A named method to run (conjoint, MaxDiff, pricing) |
| The researcher's role | Trusted route-finder | Fast, compliant executor |
| Typical methods | Mostly qualitative | Advanced quantitative and mixed methods |
With AI, your Chief Product Officer works out that conjoint analysis can help optimise your product's recommendation engine. Your Chief Design Officer realises they can create personalised paywalls based on a pairwise comparison survey that uncovers each customer segment's top unmet needs. Your CEO learns that a survey combining Gabor Granger and MaxDiff can identify which existing features have the most potential as add-on purchases.
These aren't hypotheticals. The stories from UX Researchers became a clear pattern, and OpinionX's own website attribution data shows the same thing: signups coming from LLMs have grown sharply as AI adoption spread, and LLMs are now one of OpinionX's fastest-growing sources of new researchers, driven by non-researchers using AI to discover methods they'd never heard of, then asking their in-house researchers to run those projects.
This shift in how research projects are initiated, and how your expertise is viewed, is here to stay.
More strategic research, less strategic control
There are some obvious trade-offs here for researchers:
You have less control over which methods you use.
You'll feel out of your depth more often, pushed into advanced methodologies you've never used before.
You're expected to master new techniques on the fly, regardless of their complexity or whether your current tool stack can support them.
But there are real upsides too:
You're working on more strategic research than ever before.
Your work is closer to the heart of product, pricing and growth decisions.
Your voice at the leadership table is getting louder.
When AI commoditises software development, when building product is far easier, faster and cheaper, knowing what to build, for whom, and why becomes your company's competitive advantage. And that advantage can only come from excellent user research.
This is why so much of my writing now covers pricing research methods like Gabor Granger, Conjoint Analysis, Van Westendorp and Marginal Willingness to Pay, why I keep sharing real case studies showing these methods in action, and why we keep expanding OpinionX to support more advanced research methods as fast as we can. It's why I called the newsletter The Full-Stack Researcher back in 2020.
The story of AI in UX research isn't just about keeping up. It's about staying indispensable. This shift is structural, not temporary, and embracing it is now part of what it means to succeed as a UX Researcher.
Frequently asked questions
How has AI changed the UX researcher's role? The biggest impact of AI in UX research has been to shift researchers from designing the study to executing it. Non-researchers now use AI assistants to translate strategic goals straight into named methods, like conjoint analysis or price sensitivity testing, and hand those to the researcher to run. The researcher's design expertise is increasingly bypassed, even as the work gets more strategic.
Will AI replace UX researchers? No, but it changes what they do. AI commoditises software development, which makes knowing what to build, for whom and why more valuable, not less. That insight still comes from real user research. The researchers who thrive are the ones who upskill into the advanced methods leaders now ask for.
What skills do UX researchers need in the age of AI? Increasingly, quantitative and mixed methods: conjoint analysis, MaxDiff, pricing methods like Gabor Granger and Van Westendorp, and segmentation. These let researchers speak the quantitative language of their leadership team and run the advanced studies that AI-informed colleagues now request.
Why are non-researchers requesting advanced methods like conjoint? Because AI assistants put these methods in front of non-researchers who'd never heard of them before. A CPO or CEO asks an AI how to solve a pricing or roadmap problem, gets "run a conjoint survey" as the answer, and passes that straight to their in-house researcher.
Enjoyed this post? Tens of thousands of product managers and UX researchers subscribe to our newsletter, The Full-Stack Researcher, for actionable user research advice.
About the author
Daniel Kyne is the Founder and CEO of OpinionX, the survey platform for product research. Thousands of researchers use OpinionX to measure what matters most to their most important customer segments, running advanced methods like conjoint, MaxDiff and segmentation analysis on an easy-to-use platform.
โ Try OpinionX for free today