User study vs. usability test vs. microstudy: which should you run?
Choosing a research method starts with the question you need to answer. If you want to understand why someone delayed inviting a teammate, ask about that experience. If you need to know whether they can find the invitation settings, give them a task. If you want a short answer while they are using that page, ask in the product.
Palette offers three study types for these different needs: user studies, usability tests, and microstudies. The methods can complement each other, but they are not interchangeable. This comparison helps you choose a starting point and recognize when the next question calls for a different approach.
Compare the three study types
| Method | Best starting question | Participant source in Palette | Typical evidence |
|---|---|---|---|
| User study | What do people experience, need, or expect? | Your own users or an external panel | Structured, written, or recorded answers |
| Usability test | Can people complete a meaningful task? | Your own users or an external panel | Task results, ease ratings, and follow-up feedback |
| Microstudy | What can someone tell us about this moment? | Your own users in your product | Short written or structured answers in context |
The main distinction is the kind of evidence collected. A positive opinion about a screen does not establish that a person can use it. A completed task does not explain every motivation behind using the product. A response collected on a page does not describe people who never reached that page. Keep these distinctions visible when choosing and interpreting a study.
Choose a user study for experiences and expectations
A user study gives you room to ask a sequence of questions. It is useful when the topic needs context: the person’s role, a recent event, the steps they took, and what happened afterward. In Palette, the study can combine written and recorded responses with structured formats such as select, scale, and ranking questions.
For a hypothetical invitation problem, a user study could ask recent administrators what they were trying to set up, when colleagues needed access, and what they checked before sending an invitation. Their answers may reveal that the obstacle is an organizational approval process rather than the location of a button.
This method needs participants who can speak from relevant experience and questions that do not supply the answer. Use the question-writing guide to prepare a sequence, then explore Palette user studies.
Choose a usability test for interaction with a task
A usability test is appropriate when there is an experience to use and a meaningful goal to attempt. Instead of asking whether people like the invitation screen, ask them to add a colleague with a specified level of access. The task reveals how they interpret the interface and whether they reach the intended result.
Palette lets you connect a codebase and select a branch for a sandbox version participants can use. This creates a way to examine a feature before that branch is deployed. Prepare realistic tasks and an appropriate starting state, then review task outcomes and follow-up feedback rather than relying on a single satisfaction rating.
A usability test should not be asked to prove market demand. Someone may complete the task successfully and have no reason to use the capability in their own work. Read usability testing before deployment for the setup and review process.
Choose a microstudy for a focused question in the product
A microstudy asks one or a few questions while your users remain in the product. It is useful when the setting supplies context that would be difficult to reconstruct later. For example, you might ask what information is missing while someone reviews an account settings page.
Palette lets you select pages and target custom groups. Microstudies support written responses, select questions, scales, and ranking. They are a good fit for a short question; a complex investigation may require a fuller user study. Keep the effort proportionate to the person’s reason for being on the page.
A microstudy about invitation settings can explain something about responding visitors to those settings. It cannot explain every person who abandoned onboarding earlier. For examples and interpretation limits, read the in-product microstudy guide.
Use a sequence when the uncertainty changes
You do not need to launch all three study types at once. Begin with the uncertainty that blocks the next decision. In a hypothetical onboarding investigation, a user study might reveal uncertainty about access. A usability test can then examine whether the relevant interface explains it. A later microstudy can ask a focused contextual question that remains open.
That sequence is an example rather than a fixed recipe. If the team already has strong evidence about the underlying need, starting with a task-based test may be appropriate. If people are already using the relevant screen and the unknown is narrow, a microstudy may be enough to identify the next issue to investigate.
Do not compare results from different methods as if the difference were caused by a product change. Audience, recruitment channel, question format, and study context can all influence answers. Record those conditions whenever you put findings side by side.
A short decision checklist
- Write the uncertainty as a research question.
- Decide whether you need an account of experience, evidence of task completion, or an answer in context.
- Choose participants who have the relevant role and experience.
- Check what the method and recruitment channel will leave out.
- Pilot the study, then reserve time to inspect the source responses.
For the audience choice, see recruiting your users or a panel. For a recurring practice, read continuous user research for SaaS teams. Palette’s agentic research workflow can help turn the selected question into a study while you keep responsibility for its scope and interpretation.