Concept Test Plan

Use this when the team has an idea or concept that needs validation before committing to build. A concept test sits between generative research ("what problems exist?") and usability testing ("can people use it?") -- it answers "does this direction resonate?" Produces a complete concept test plan: hypothesis, stimulus design, evaluation criteria, discussion guide, and logistics.

Related skills: Upstream: /assumption-map identifies which assumptions need concept validation. Pairs with /screener-design for participant recruitment. Downstream: feeds into /research-synthesize for analysis. If the concept passes, move to /usability-test-plan for evaluative testing. Part of the discovery-sprint recipe.

Process

Step 1: Gather inputs

Ask the user to provide:

1. The concept -- what idea, feature, or direction are you testing? Describe it in plain language, not product jargon.
2. What you're trying to learn -- what decision does this test inform? (e.g., "Should we invest a sprint in building this?" or "Which of three directions should we pursue?")
3. Assumptions being tested -- what has to be true for this concept to succeed? (Link to /assumption-map output if available.)
4. User segment -- who should evaluate this concept? What context do they need to have?
5. Constraints -- timeline, number of participants, remote vs. in-person, what stimulus materials exist (or need to be created).

Step 2: Define hypothesis and evaluation criteria

Frame the test around a falsifiable hypothesis:

## Concept Test Plan -- (Concept name, date)

### Hypothesis
We believe that (target users) will (expected reaction) when presented with (concept) because (rationale). We'll know this is true when (measurable signal).

### Evaluation criteria

| Signal | Positive indicator | Negative indicator | How captured |
|---|---|---|---|
| Comprehension | Participant explains concept in own words accurately | Participant misunderstands purpose or confuses with existing solutions | Think-aloud during concept review |
| Relevance | "I would use this" or describes personal use case | "I don't see why I'd need this" or can't name a situation | Direct question + follow-up probe |
| Value | Ranks concept above current workaround | Prefers status quo or competing solution | Comparison question |
| Credibility | Believes the concept would work as described | Expresses doubt about feasibility or trustworthiness | Probe after initial reaction |
| Willingness to act | Would sign up, pay, or switch | "Maybe someday" or conditions that will never be met | Commitment question |

Evaluation criteria rules:

• Include at least 3 signals. Comprehension, relevance, and value are non-negotiable. Add credibility, willingness to act, or domain-specific signals as needed.
• Positive and negative indicators must be specific observable behaviors or quotes, not interpretations.
• "Participants like it" is not a valid signal. What they say, do, or choose is.

Step 3: Design the stimulus

Choose the right fidelity for the concept being tested:

### Stimulus design

**Format:** (One of the options below)

| Fidelity level | When to use | Examples | Pros | Cons |
|---|---|---|---|---|
| Verbal description | Very early concepts, abstract ideas | Elevator pitch, scenario narrative | Fast, no design needed | Easy to misunderstand, leading |
| Storyboard | User journey concepts, multi-step flows | 4-6 panel comic, annotated wireflow | Shows context and sequence | Requires illustration effort |
| Static mockup | UI concepts, feature ideas | Figma screens, annotated screenshots | Concrete, testable | Can fixate on visual details |
| Interactive prototype | Near-build concepts, flow validation | Clickable Figma, coded prototype | Realistic interaction | Expensive to create, hard to change |
| Video / animation | Complex interactions, system concepts | Screen recording with narration, motion prototype | Shows dynamics | One-way, no exploration |
| Concept card | Comparing multiple concepts | Name + tagline + 3 bullet value props per concept | Easy comparison | Thin, may not convey enough |

**Stimulus description:**
- (What will be shown to participants)
- (What level of polish -- rough/medium/polished)
- (What context or framing will accompany it)
- (What will NOT be shown -- scope boundaries)

**Stimulus review checklist:**
- [ ] Does the stimulus convey the concept without the moderator explaining it?
- [ ] Is the fidelity appropriate? (Not so rough that participants can't react, not so polished that they think it's finished)
- [ ] Does it avoid leading the participant toward a specific reaction?
- [ ] If comparing concepts, are they presented at equal fidelity?

Step 4: Build the discussion guide

### Discussion guide

**Before stimulus (5-10 min)**
1. Welcome, consent, think-aloud instruction
2. Context questions about their current workflow:
   - "Tell me about the last time you (relevant activity)."
   - "What do you currently use for (problem area)?"
   - "What's the most frustrating part of (current process)?"

**Stimulus presentation (5-10 min)**
3. Present the concept. Say: "I'm going to show you something we're exploring. It's not finished -- we're looking for your honest reaction."
4. Pause. Let them react. Note their first words -- these are the most honest signal.
5. Comprehension check: "In your own words, what would this do for you?"

**Evaluation probes (15-20 min)**
6. Relevance: "When would you use this? Describe a specific situation."
7. Value: "How does this compare to what you do now? Better, worse, or different?"
8. Concerns: "What worries you about this?" (Not "do you have concerns" -- assume they do.)
9. Missing: "What's missing that would make this useful to you?"
10. Priority: "If you had to choose between this and (alternative/status quo), which would you pick? Why?"

**Commitment test (5 min)**
11. (Choose one or more):
    - "If this existed today, would you try it?" (then probe: "What would you try first?")
    - "Would you recommend this to a colleague? Who specifically?"
    - "What would you give up to get this?" (time, money, switching costs)

**Debrief (5 min)**
12. "What's the one thing you'd change about this concept?"
13. "Anything else we should know?"

Discussion guide rules:

• Context questions come BEFORE showing the concept. Never anchor participants on your idea before understanding their world.
• The first reaction after stimulus presentation is the most valuable data point. Do not talk over it or prompt too quickly.
• "Would you use this?" is a weak question (most people say yes to be polite). "When would you use this?" and "What would you give up?" are stronger.
• If testing multiple concepts, randomize presentation order across participants to control for order effects.

Step 5: Plan logistics

### Logistics

| Item | Detail |
|---|---|
| Participants | (6-8 typical for concept tests -- more than usability, fewer than survey) |
| Session length | (30-45 min per session) |
| Format | (1:1 interview with stimulus -- not focus group) |
| Setting | (Remote / in-person) |
| Recruitment | (Use /screener-design) |
| Stimulus prep | (Who creates it, by when) |
| Schedule | (Date range, sessions per day -- max 4 to avoid moderator fatigue) |
| Comparison design | (If testing multiple concepts: within-subject rotation or between-subjects) |
| Recording | (Video + audio + notes -- confirm consent) |

Step 6: Review and validate

Ask the user:

• Is the hypothesis falsifiable? What result would kill this concept?
• Is the stimulus fidelity right? Too rough and participants can't react; too polished and they think it's decided.
• Are you testing the concept or the execution? (Concept tests should answer "is this worth building?" not "is this the right UI.")
• If you're comparing concepts, are they genuinely different directions or minor variations? (Variations should be A/B tested, not concept tested.)
• Who needs to see results, and what format? (Consider /research-readout for stakeholder presentation.)

Output location

Present the concept test plan as formatted text in the conversation. The user copies it into their research documentation or shares with the team.