Stats
Actions
Tags
From grimoire
Structures user research studies: defines research questions, selects methods, drafts participant screeners, and sets timelines before recruiting.
How this skill is triggered — by the user, by Claude, or both
Slash command
/grimoire:design-research-planThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Write a one-page research plan that aligns the team on what questions the study will answer, which method fits those questions, and who to recruit — before a single participant is contacted.
Write a one-page research plan that aligns the team on what questions the study will answer, which method fits those questions, and who to recruit — before a single participant is contacted.
Adopted by: Google Ventures mandates a written research plan before every sprint research day; IDEO's HCD Toolkit requires a "learning goals" document as the first step in all field research; Nielsen Norman Group's UX research training treats the research plan as the non-negotiable foundation of any credible study Impact: NNG research shows that teams without a written plan collect data that answers the wrong question in the majority of studies — the most expensive research mistake because it is only discovered after sessions are complete and cannot be fixed retrospectively; a plan reviewed by stakeholders before recruiting eliminates post-study scope disputes and prevents "we should have asked about X" regret Why best: Ad-hoc research — starting with recruiting before aligning on questions — produces data that is hard to act on; different team members leave sessions with different interpretations of what was learned; a written plan is the contract between the research team and stakeholders that prevents both under-scoping (missing key questions) and over-scoping (trying to answer everything in one study)
Sources: NNG "UX Research Cheat Sheet" (Nielsen Norman Group, 2017); IDEO "The Field Guide to Human-Centered Design" (IDEO.org, 2015); Goodman, Kuniavsky & Moed "Observing the User Experience" (Morgan Kaufmann, 2012) Ch. 3
One sentence. Not a business question ("should we build this?") and not a solution ("how do we improve checkout?") — a behavioral or attitudinal question about users:
What are the primary reasons users abandon the checkout flow before entering payment information?
Gates:
If the team cannot agree on one research question, run a 20-minute alignment session to prioritize — multiple competing questions signal a scope problem, not a research plan problem.
| Question type | Method |
|---|---|
| Why do users behave a certain way? | User interviews |
| Can users complete tasks on the interface? | Usability testing |
| What do users do in their natural environment? | Contextual inquiry / diary study |
| How do users mentally organize a topic? | Card sorting |
| What are users' attitudes at scale? | Survey |
| Which of two designs performs better? | A/B test |
One method per study. If multiple question types are present, split into multiple studies or prioritize the highest-risk unknown.
Write a screener with three sections:
Qualifying criteria — must-haves that define the target user:
Disqualifying criteria — exclude anyone who would skew results:
Desired mix — diversity factors relevant to your question:
| Method | Recommended count |
|---|---|
| Interviews | 5–8 per distinct user segment |
| Usability testing (formative) | 5 per segment |
| Usability testing (summative) | 20+ for statistical validity |
| Card sorting | 15–20 |
| Survey | 100+ for quantitative analysis |
| Contextual inquiry | 5–8 |
More participants are not always better for qualitative methods — the 5-user rule (Nielsen, 1993) applies to formative usability testing; diminishing returns set in quickly for interviews.
Week 1: Finalize plan + screener → stakeholder review
Week 2: Recruit participants (allow 5–7 business days minimum)
Week 3: Run sessions (1–2 per day max to preserve researcher quality)
Week 4: Synthesize findings → readout
Buffer for: no-shows (recruit 20% more than needed), scheduling conflicts, pilot session rescheduling.
Share the one-page plan with PM, design lead, and any stakeholder who will act on findings. Require explicit sign-off before recruiting begins. The questions they add or remove at this stage cost nothing; the questions they add after sessions end cost another full round of research.
npx claudepluginhub jeffreytse/grimoire --plugin grimoireGenerates a complete UX research plan including objectives, methodology, screener questions, discussion guide, and synthesis framework. Activates on research or usability testing requests.
Plan and execute user research: frame questions, choose methods, recruit participants, conduct interviews, synthesize findings, and drive product decisions.
Plans user research by identifying knowledge gaps, framing observable questions, selecting methods like usability testing and interviews, and ensuring participant inclusion.