dscout 探讨 AI 改变 UX 实践后保持相关性的十种方法

AI continues to widen the path of UX research, making it increasingly accessible to more people. And as a UX practitioner, sometimes it can feel hard to catch up—or even understand where your place is anymore in the UX research process. 

Stevie Vanderwiel (Dscout’s Senior Manager of Brand and Growth Marketing), sat down with three experts to discuss:

• How UX research roles continue to evolve
• What still remains the same
• What you can uniquely bring to the table when everyone is more capable of generating insights

Panelists included:

• Saeideh Bakhshi, Quant UX Researcher at OpenAI
• Beth Lingard, Product Strategist and Advisor at Lingard Strategy Studio and former Research Director at Meta
• Andy Warr, Product Manager at Dropbox

This article is based off the webinar When Everyone Can Generate Insights, What Matters Most? You can watch it in its entirety here. 

What has (and hasn’t) changed with AI in the mix

The bottleneck has moved 

Years ago, cloud computing made collecting massive amounts of data possible, which moved the hard part from logging numbers to hiring people who could make sense of them. AI is doing something similar today by taking away the pain of generating quick insights. 

The bottleneck isn't getting the information; it's what happens at the end of the line. The work is now about verification, double-checking quality, and synthesizing everything that comes in.

More insights also means more noise 

When data can be pulled from almost anywhere, having more insights around can actually create a lot of extra noise. A key part of the job now is helping teams figure out which insights are worth paying attention to, and whether we're even asking the right questions.

UXRs still need to center the customer voice 

Sorting quality information from noise still takes time, care, and reflection. Even as data collection gets easier, researchers need to make sure the customers’ voice doesn't get lost in all the chatter.

“UXRs need to make sure that customer voice is not lost in a sea of data.”

- Andy Warr, Product Manager at Dropbox 

10 ways to level up your UX practice amidst the changes

1. Provide value when thinking deeply about the user 

Try leveraging your position to look across feature silos and focus on the whole user journey. While product managers often focus on specific feature areas, you can bring value by looking across team lines to understand end-to-end human needs. 

Step back, ask fundamental questions about what problem is actually being solved, and help your team scope what really needs to be studied. That broader view helps you spot experience gaps that individual product teams might overlook.

2. Offer expertise during the interpretation phase 

Focus your energy on interpreting what findings mean—especially when non-researchers use AI to pull quick summaries. When self-serve tools spit out multiple plausible explanations for the same data, that’s also a great opportunity to step in and untangle conflicting patterns. 

Doing so also means you can act as an expert reviewer who brings much-needed context to the table. Your interpretation layer helps decision-makers feel confident when the stakes are high.

3. Focus on real customers and educate how to make sense of data at scale 

Keep your organization anchored in real, living customers (especially as AI tools and synthetic users make scaled data feel tempting). Remember that larger sample sizes don't automatically guarantee better decisions if teams don't know how to interpret what they're seeing. You can also educate your cross-functional partners on how to make sense of scaled data responsibly. This helps your team avoid a false sense of security and ensures everyone relies on verified, human insights.

4. Bridge teams and think more fluidly 

As more people conduct research, act as the connective tissue that links different cross-functional teams across standard org charts. You can follow the user's actual journey as it crosses team lines, and use those learnings to bring disparate groups together around human outcomes. Thinking fluidly helps you make sure critical parts of the user experience don't fall between the cracks of your company's structure.

5. Shift your focus from execution alone

• Lean into your expertise: Double down on your deep domain knowledge and methodology, as these become your strongest differentiators when basic tasks get automated. Position yourself as the go-to expert for messy, complex, or high-stakes questions where self-serve tools fall short.
• Make your range more versatile: Don't stop at handing off static reports to designers or PMs. Expand your operational range by taking user feedback and directly building or prototyping ideas to test with users next.
• Be adaptable with process and workflows: Test out AI tools and reimagine your standard research steps instead of clinging rigidly to traditional timelines. Focus on core research principles while letting go of procedural constraints to keep your practice agile.

6. Know when to take a beat 

Remember that speed isn't a strategy, and give your team permission to slow down when needed. Ask simple, grounding questions like "Who are we solving this for?" to encourage everyone to pause and reflect. Taking a beat keeps high-stakes decisions from being rushed just because automated tools generate quick answers.

“Some of the most important step-change innovation comes from slowing down and thinking. Let things breathe a little bit.”

- Beth Lingard, Product Strategist and Advisor at Lingard Strategy Studio and former Research Director at Meta

7. Always prioritize peer review 

Borrow a page from software developers and set up peer review habits for your research work. Ask a colleague to review your study designs and survey mechanics before launching them to boost overall quality through feedback. As AI takes over heavy execution tasks, use that extra time to establish quality checks across your company.

8. Provide value during the scoping and interpretation phase 

Shift your energy away from basic data collection and sorting, which AI can easily automate. Instead, focus on the bookends of the research process: framing the right questions during scoping and drawing actionable meaning during interpretation. Owning these crucial stages allows you to translate raw research into sound product choices you can confidently stand behind.

9. Define a higher tier for yourself 

AI has already automated and streamlined many repetitive research tasks. Instead of spending energy defending traditional role boundaries, lean into current shifts to step into higher-level strategic work. Clearing out tactical overhead frees up room for you to shape bigger business and product decisions.

10. Embrace the ambiguity 

It’s easy to feel fear or even existential dread about these big changes. Try to lean into your human-centered training and approach these big changes with curiosity rather than fear. You can use this shifting landscape as an opportunity to acquire new skills or explore adjacent fields like product management, design, or data science.

“Our moat was never about the UX research, the way it is defined through tools, through methods, through approaches.

Our moat is about understanding people and bringing that understanding to the product. That is increasingly more and more important.” 

- Saeideh Bakhshi, Quant UX Researcher at OpenAI

Wrapping it up

The continued accessibility of UX research practices may initially feel like a threat to your work. But your experience and expertise is more important than ever to QA, govern, safeguard, advise, and facilitate the work across your company. 

Use this change as an opportunity to:

• Continue learning with a curious mind
• Provide more guidance and expertise at the beginning of studies and interpretation phase of insights
• Keep the customer voice at the center and dial down the noise 
• Slow down and listen to your inner compass on what approaches, data, and insights are the most impactful