Summary
A Research Plan forces clarity on goals, aligns stakeholders, and ensures every step serves core questions. Key elements include research goals and questions, participant definition, hypotheses, and methodology. The plan should be a living document updated throughout the project. Before any empirical study, consider starting with a heuristic evaluation to clean up obvious issues first.
Good research does not happen by accident; it is the result of careful, deliberate planning. Before you write a single interview question or schedule a participant, you need a Research Plan.
This document is your blueprint. It forces you to clarify your thinking, align with stakeholders, and ensure every step is designed to answer your core questions. It is the single most important tool for avoiding unfocused, low-impact research, and for enabling good communication throughout the project.
Step 0: The Heuristic Evaluation
Before testing with users, always perform an expert review. Teams often skip this, but it is the most expensive mistake you can make. Use Jakob Nielsen's 10 Usability Heuristics [1] to catch obvious issues first. Save your participant budget for the complex problems that experts cannot find.
The Golden Rules of Question Design
When designing questions to ask participants, your job is to translate high-level research goals into concrete, answerable questions. These three rules will keep you out of trouble.
Rule 1: Favor Behavior Over Opinion
Do not ask "Are you interested in fitness?" (Opinion). Ask "How many times did you work out last week?" (Behavior). Opinions are aspirational; behavior is reality.
- Weak: "Are you interested in fitness?"
- Strong: "How many times did you work out last week?"
The most reliable insights come from what people have done, not what they say they will do.
Rule 2: Avoid Hypotheticals
While hypothetical questions can cautiously explore ideas in early generative research, they should never be treated as reliable predictions of future behavior.
Rule 3: Ensure Categories Are MECE
For multiple choice questions, answers must be Mutually Exclusive and Collectively Exhaustive (MECE). A user should never fit into two categories or no categories.
- Broken: "Age 20-30" and "Age 30-40" (30-year-olds fit both)
- Correct: "Age 20-29" and "Age 30-39" (no overlap)
Always include "Other (please specify)" or "I don't know" options to ensure exhaustiveness.
Bonus: Probe, Don't Lead
Frame questions neutrally to avoid influencing the answer.
- Leading: "Was that checkout process easy?"
- Neutral: "How was your experience with the checkout process?"
Elements of a Research Plan
A good research plan does not need to be a 50-page thesis. It can be a simple, clear document that answers foundational questions in logical order: Why are we doing this? Who is it for? What do we expect to find? How will we get it done?
1. Research Goals and Questions
This is the most important section. Define the problem you are trying to solve and the business decisions the research will inform.
- Weak goal: "We want to get feedback on the new checkout flow."
- Strong goal: "We need to understand the primary points of friction and abandonment in our new checkout prototype for first-time mobile users, so the design team can prioritize fixes before the next development sprint."
You must translate high-level goals into concrete, answerable questions:
- "At this step, what are your thoughts?" (probing for in-the-moment cognition)
- "Was there anything on this page that you found unclear or surprising?" (specific post-experience question)
2. Participants
Based on your goals, define the specific characteristics of the people you need to study, going beyond demographics to include relevant behaviors, attitudes, and experiences.
- Gaming example: "Participants who play strategy games on PC for at least 8 hours per week and have experience with both real-time and turn-based genres."
- Healthcare example: "Participants diagnosed with Type 2 diabetes within the last 2 years who currently use a mobile app to track blood glucose levels."
3. Hypotheses
You are rarely starting from a blank slate. Based on previous research or analytics, you should have informed guesses about what you expect to find.
Stating these as hypotheses is powerful: it forces you to look for evidence that either supports or refutes your existing beliefs, preventing you from simply finding what you want to find.
- "We believe users will struggle to find the option to use a different billing address because the link is not prominent enough."
- "We believe users will express concern about the lack of familiar payment logos on the final confirmation screen."
4. Methodology, Environment, and Materials
Outline the specific method and approach (qualitative, quantitative, or mixed).
Research Environment is a critical decision involving a fundamental trade-off between control and ecological validity (realism):
| Environment | Best For | Trade-offs |
|---|---|---|
| Moderated In-Person | IDIs, hardware testing, complex software, sensitive topics | Rich data, but limited geographic reach |
| Moderated Remote | Any study requiring live interaction | Geographic reach, cost-effective, but dependent on participant's tech setup |
| Unmoderated Remote | Large-scale surveys, card sorting, tree testing, unmoderated usability tests | Speed and scale, but no live probing or clarification |
The Plan as a Living Document
It is a common mistake to treat the research plan as a static document that gets filed away once fieldwork begins. You write the plan, get sign-off, and never look at it again. This is a missed opportunity.
The most effective research plans are living documents. They get updated throughout planning and execution. They accumulate context that becomes essential later.
The Changelog
Your plan should include a dedicated section for ongoing documentation. Think of it as a changelog or field journal. Every time something significant happens, you write it down.
What to document:
| Category | Examples |
|---|---|
| Plan History | Stakeholder discussions that changed the scope. Decisions to add or remove research questions. Methodology pivots based on new information. |
| Technical Issues | Recording failures. Screen sharing problems. Prototype bugs discovered mid-session. |
| Protocol Deviations | Sessions that ran long or short. Questions you forgot to ask. Tasks you had to skip due to time. |
| Participant Anomalies | No-shows and replacements. Participants who did not match screening criteria. Interruptions or distractions during sessions. |
| Unexpected Observations | Patterns emerging earlier than expected. Behaviors that contradict your hypotheses. Quotes worth flagging immediately. |
Why This Matters
This habit of documentation pays off in three ways.
Single source of truth. When a stakeholder asks "why did we change the screener halfway through?" you have the answer. When a team member joins late, they can read the changelog to understand how the project evolved.
Context for analysis. Three months from now, when you are analyzing data and find an outlier, the changelog tells you that Participant 7 had a crying child in the background for the first 15 minutes. That context changes how you interpret their task times.
Honest interpretation. Research rarely goes exactly as planned. Documenting deviations forces you to account for them in your conclusions. You cannot claim clean findings if your field notes reveal messy execution. This transparency strengthens, rather than weakens, the credibility of your work.
Key Artifacts
Before any session, finalize your discussion/test guide. This script ensures consistency and should include:
- Introduction and informed consent
- Warm-up questions
- Core interview or survey questions
- For UX Tests:
- Task scenarios (framed around user goals, not system functions)
- Post-task questions
- Post-session wrap-up
What This Means for Practice
The research plan is your contract with stakeholders. It documents:
- What you will do and what you will not do
- What kind of conclusions the research can support
- What decisions it will inform
Get alignment on this before starting fieldwork. A plan that evolves through discussion is far more valuable than one that sits unread in a folder.
Without a plan, research drifts. With a plan, every activity has purpose.