Moderated vs. Unmoderated UX Research: Pros, Cons, and How AI Fits In
SUMMARY
Moderated and unmoderated research can both deliver valuable user insights, but they differ in depth, speed, cost, scale, and complexity. Here’s how to weigh the pros and cons, where AI can support either approach, and the best practices that help UX research succeed regardless of methodology.
Table of Contents
Choosing the Right Research Approach Starts with Understanding the Trade-offs
At Sandstorm, we have conducted more than 5,000 hours of UX research studies, and one thing is clear: there is no single way to conduct effective user research.
The right methodology depends on what you need to learn, who you need to hear from, the complexity of the questions you are asking, and practical considerations such as timeline and budget.
Now, there is another consideration: AI.
AI is changing how researchers analyze and synthesize findings, creating new efficiencies across both moderated and unmoderated approaches. But technology is only part of the equation.
Understanding the strengths and limitations of each approach is the first step toward choosing the right one.
Option 1: Moderated User Research
In a moderated study, a researcher meets directly with participants by video or in person and guides them through questions and activities.
Because the researcher is part of the conversation, they can observe behavior, ask follow-up questions, explore unexpected findings, and adjust the session in real time.
Pros of Moderated Research
- Go deeper
A researcher can ask “why?” and explore the thinking behind a participant’s actions rather than relying solely on what they did. - Follow unexpected findings
An experienced moderator can recognize when a comment, hesitation, or behavior deserves further exploration and adjust the conversation accordingly. - Combine multiple research questions and methodologies
A moderated session can move between different types of research while the moderator provides context and guides participants through each transition. One session might include a tree test, brand perception questions, and exploration of a prototype, helping teams learn more from each participant without requiring separate rounds of recruitment. - Explore complex experiences
Complicated workflows, early concepts, sensitive subjects, and nuanced decisions often benefit from conversation and context. - Observe more than task completion
Hesitation, confusion, tone, workarounds, questions, and reactions can reveal insights beyond whether someone completed a task. - Adapt in real time
If a participant misunderstands an instruction or encounters a technical problem, the moderator can adjust while preserving the intent of the research.
Cons of Moderated Research
- More time-intensive
Individual sessions need to be recruited, scheduled, conducted, reviewed, and analyzed. - Higher research cost
More researcher time generally means a greater investment. - Scheduling can limit participation
Participants need to be available at a specific time, which can make certain audiences harder to reach. - Typically involves fewer participants
Because each session requires researcher time, sample sizes are often smaller.
Option 2: Unmoderated User Research
In an unmoderated study, participants complete activities independently using a research platform such as Lyssna and Optimal Workshop. There is no researcher present during the individual session.
The study has to do more of the work on its own, but removing the need for a researcher to attend every session creates advantages in speed, flexibility, cost, and scale.
Pros of Unmoderated Research
- Faster to execute
Participants can complete research on their own schedules without coordinating individual sessions. - More cost-efficient
Less researcher time is required for each participant. - Easier to scale
Teams can often gather feedback from more participants in a shorter period of time. - More flexible for participants
People can participate when it is convenient rather than committing to a scheduled interview. - Consistent study experience
Every participant receives the same tasks and instructions. - Effective for focused questions
Findability, navigation, comprehension, preference, and straightforward task completion can be particularly well suited to unmoderated testing.
Cons of Unmoderated Research
- You can lose context and nuance
What someone did may be unclear. What they were thinking when they did it can be harder to determine. While it's possible to record screen and audio, this can increase friction of user permissions needing a computer and Chrome-based browser. - More questions can create more frustration
Unmoderated studies benefit from being short and focused. Too many questions or tasks can make the platform experience slower and the study feel longer, increasing the risk of frustration or abandonment. - Technical friction can cost participants
Page-loading, prototype limitations, browser permissions, and recording requirements can cause people to abandon a study. - Instructions need to stand on their own
If a participant misunderstands a task, there is no moderator there to clarify it.
Unmoderated Does Not Mean Unattended
Because there is no researcher in the room, unmoderated research can appear deceptively simple: build the study, send the link, and wait for the results.
In practice, it requires careful planning.
As a general guideline, we recommend limiting an unmoderated usability study to approximately five or six tasks.
There is another experience worth testing before launch: the research itself.
Click the invitation. Open the platform. Try the prototype. Test different browsers and devices. See what happens if you decline microphone access.
Recording can provide valuable context, but requiring audio or screen recording may also create a barrier. When possible, make recording opt-in. If it is essential to the research, communicate the requirement before participants begin.
Good UX research should also be good UX.
Prototype or Live Website? Consider the Experience
Prototypes are incredibly valuable when an experience does not exist yet or when you need to isolate and evaluate a particular concept. They can be especially useful for testing navigation concepts, such as a new mega menu, before development is complete.
But prototypes introduce artificial boundaries. Links may not work everywhere participants expect, interactions may behave differently, and participants can encounter limitations unrelated to the design you intended to evaluate.
When the website is already available, testing the live experience can reduce some of that friction.
That creates another opportunity: post-launch research.
Consider a focused unmoderated study two to four weeks after launch to validate key decisions and identify early opportunities for optimization. It does not need to become another major research initiative. A targeted study of approximately 10 questions paired with a concise executive summary can provide meaningful insight while the new experience is still fresh.
Launch does not need to be the finish line for UX research.
Recruitment is Part of the Research Strategy
Even a well-designed study needs the right participants at the right time.
A few ways to improve participation:
- Give research its own moment
A dedicated research email is more likely to get noticed than a link competing with other content in a newsletter. - Check the calendar
Consider other client communications, annual meetings, conferences, holidays, and events that may compete for your audience’s attention. If you are researching residents, patients, members, or another specific audience, think about what may be happening in their lives when you are asking for their time. - Go where your audience already is
Annual meetings, conferences, and other gatherings can be great opportunities for research because participants are already engaged and accessible. Consider short interviews, intercept studies, prototype testing, or other focused activities. - Make the invitation meaningful
Tell people why their perspective matters, give them an inside look at what you are working on, and explain how their feedback will help shape what comes next. - Consider incentives
Compensation can help drive participation, particularly when you are asking for significant time, effort, or specialized knowledge.
Plan recruitment early and consider:
- Who needs to participate
- How and where you can reach them
- What else is competing for their attention
- Whether an upcoming meeting, conference, or event creates a research opportunity
- Whether dedicated email outreach is available
- Whether incentives are appropriate
- What approvals are required
Recruitment strategy is research strategy.
Where Does AI Fit Into UX Research?
Across both approaches.
AI is increasingly part of the researcher’s toolkit rather than a methodology unto itself.
Whether sessions are moderated or unmoderated, AI can help researchers:
- Transcribe participant sessions
- Summarize individual sessions or responses
- Synthesize feedback across participants
- Organize large volumes of qualitative data
- Surface recurring themes and patterns for further investigation
- Accelerate portions of the analysis process
That can make research considerably more efficient, especially when a study generates hours of recordings, transcripts, or open-ended responses.
But efficiency and interpretation are different things.
AI might surface a recurring complaint about a navigation label. A researcher still needs to understand the context, determine whether the pattern is meaningful, consider who experienced the issue, compare it with behavioral observations and other evidence, and determine what should happen next.
AI can help us get through the data faster. Human expertise helps determine what the data means.
Start With What You Need to Learn
Moderated and unmoderated research give us different ways to listen to users.
Moderated research provides conversation, context, adaptability, and the ability to combine more complex research activities. Unmoderated research provides speed, flexibility, scale, and cost efficiency. AI can make aspects of both approaches faster and more manageable.
Sometimes one methodology makes sense. Sometimes the other does. And sometimes the strongest research plan uses both at different points in the process.
At Sandstorm, we help clients determine what they need to learn, choose the right approach, reduce subjectivity, and turn research into actionable recommendations.
Rather than starting with the tool, start with the question:
What do we need to learn, and which approach will give us the evidence we need to make a better decision?
The tools will continue to change. The fundamentals of good UX research will not.
Need help choosing the right UX research approach? Sandstorm can help you find the best fit for your goals and turn user insights into actionable recommendations. Let's talk.
