95% of new products fail. That number comes from Harvard Business School professor Clayton Christensen and it has been haunting product teams ever since. The reason is almost never a bad idea. It is usually a good idea to ask about it before spending the budget. That is exactly what concept testing questions are designed to prevent. They are the questions you ask real people before you build, not after you have already built the thing and realized nobody wanted it. If you are a product manager, UX researcher or founder who wants to avoid being a cautionary tale at someone else’s keynote, this guide is for you.
TL;DR
- What are concept testing questions? Targeted questions that measure whether an idea is worth building before you build it.
- Why do most concept tests fail? Teams ask for purchase intent before comprehension and bake in bias from the start.
- 5 question types to know: Comprehension, value perception, purchase intent, comparative and open-ended.
- Stage-by-stage framework: Different stages need different questions. Discovery is not the same as late validation.
- 16 questions with examples: Real questions mapped to stage, format and red-flag signals.
- How to analyze results: High interest plus low purchase intent usually means novelty trap, not product-market fit.
What Are Concept Testing Questions?
Concept testing questions are targeted questions used to measure the potential success of an idea, product feature or service before any serious development begins. Think of them as a first date for your product. You want to know if there is any chemistry before you move in together.
They are not the same as usability testing questions, which test how people interact with something that already exists. And they are different from product testing, which evaluates a finished or near-finished prototype. Concept testing happens earlier, at the drawing board stage, when the concept is still an idea with a rough sketch and a lot of hope.
When Does Concept Testing Apply?
- Launching a new product in a market you have never entered
- Testing a feature pivot before committing engineering resources
- Validating a brand refresh or new messaging angle
- Choosing between two product directions before the roadmap locks in
Why Most Concept Tests Fail Before They Start
Before we get to the good questions, let us talk about the bad ones. Most concept tests underperform not because the product idea is terrible but because the questions are designed to confirm what the team already believes.
The Purchase Intent First Problem
Every article on concept testing will tell you to ask “How likely are you to purchase this product?” right out of the gate. There is a problem with that. If someone has not yet understood what the product does or why it exists, their purchase intent answer is based on nothing. You are asking them to vote before they have read the ballot.
The correct order is: comprehension first, evaluation second, purchase intent third. Reverse this and you inflate your scores.
Question Order Bias
When purchase intent comes before comprehension, respondents anchor to their initial gut reaction and then rationalize it. This is called anchoring bias and it is especially dangerous in concept testing because it gives you confident-looking data that is actually built on a misunderstanding.
What Bad Concept Testing Questions Look Like
Recognizing a bad question is just as important as writing a good one. Here are the main offenders:
Leading Questions
These push the respondent toward the answer you want. “Do you agree that this innovative new feature would save you time?” is not a question. It is a compliment in disguise.
Double-Barreled Questions
Asking two things in one question is a classic mistake. “Is this product useful and affordable?” gets you one answer for two very different things. Split them.
Internal Bias in Question Design
When the team that built the concept also writes the survey questions, bias sneaks in quietly. Words like “innovative,” “seamless,” and “efficient” in a question description pre-sell the concept before a single question is asked.
5 Types of Concept Testing Questions (and When to Use Each)
A strong concept test does not just ask if people like your idea. It digs into how they understand it, whether it fits their life and what it would take for them to actually pay for it.
1. Comprehension Questions
These check whether respondents understand what your concept actually is. If they cannot describe it back to you, nothing else you measure will be accurate. Use these at the start of every concept test, no exceptions.
“In your own words, what does this product do?”
2. Value Perception Questions
Once they understand it, ask if it matters to them. A clear concept that solves a problem nobody has is still a bad product.
“What problem does this product solve for you, if any?”
3. Purchase Intent and Likelihood Questions
These are the numbers your stakeholders will ask about. Ask them third, not first. A 5-point Likert scale works well here.
“How likely are you to purchase this if it were available today?”
4. Comparative and Differentiation Questions
These tell you whether your concept stands out in a crowded market or gets lost in it.
“How does this compare to the solution you currently use for this problem?”
5. Open-Ended and Qualitative Questions
These are where the real insights hide. Numbers tell you what people think. Open-ended answers tell you why.
“What, if anything, would stop you from using this product?”
Quick Format Guide
The format of the question changes what kind of data you get:
- Likert scale (1 to 5 or 1 to 7): Great for purchase intent and satisfaction
- Ranking: Use when comparing multiple features or product concepts
- Rating: Similar to ranking but answers do not have to be mutually exclusive
- Open field: Essential for qualitative depth; harder to analyze at scale
Stage-by-Stage Question Framework
Here is the gap that no other article talks about. Not all concept testing questions belong at every stage of development. Using late-validation questions at the discovery stage is like asking someone to sign a lease before they have seen the apartment. According to research compiled by Highlight, companies that implement comprehensive testing protocols report failure rate reductions of 30 to 50 percent. The key is making the testing stage-specific, not a one-time survey.
Stage 1: Discovery (Does the Pain Exist?)
At this stage, you do not pitch the concept at all. You are just trying to confirm the problem is real and frequent enough to matter.
- “How do you currently solve [the problem your concept addresses]?”
- “How satisfied are you with that solution on a scale of 1 to 5?”
- “How often does this problem come up for you in a typical week?”
Ideal question count at this stage: 5 to 8 questions, mostly open-ended or Likert scale.
Stage 2: Early Concept (Does the Solution Land?)
Now you introduce the concept. Your questions focus on comprehension and value, not buying.
- “In your own words, what does this product do?”
- “What do you find most appealing about this concept?”
- “What is missing or unclear?”
Ideal question count: 8 to 12 questions. Mix of open-ended and scaled questions.
Stage 3: Late Validation (Will They Commit?)
Only at this stage should you lean into purchase intent, pricing and NPS questions. The concept is refined enough for those answers to mean something.
- “How likely are you to purchase this product if it were available at [price point]?”
- “How likely are you to recommend this to a colleague or friend? (NPS scale)”
- “Which pricing option would you consider fair for this product?”
Ideal question count: 10 to 15 questions, with a good mix of quantitative and one or two open-ended follow-ups.
16 Concept Testing Questions to Ask (With Examples)
Below are 16 product concept testing questions organized by stage. Each one includes a note on whether it works better in a moderated interview or an unmoderated survey, plus a red-flag pattern to watch for in the response data.
Comprehension Questions
Q1. “In your own words, what does this product do?”
Format: Open-ended. Works in both interviews and surveys. Red flag: If more than 30 percent of respondents describe it incorrectly, the concept needs a clearer explanation before you test anything else.
Q2. “Who do you think this product is designed for?”
Format: Open-ended. Best in the interview. Red flag: If they describe a customer segment you were not targeting, you have a positioning problem.
Value Perception Questions
Q3. “What problem does this product solve for you, if any?”
Format: Open-ended. Works in both. Red flag: “I already have a solution for this” as a common theme means you have a differentiation gap, not a product gap.
Q4. “On a scale of 1 to 5, how relevant is this to your daily life or work?”
Format: Likert scale. Survey-friendly. Red flag: Scores of 3 or below across most respondents signal low product-market fit.
Q5. “What would you have to give up to use this product?”
Format: Open-ended. Best in the interview. Red flag: If the switching cost sounds too high, the value proposition needs to be stronger.
Purchase Intent Questions
Q6. “How likely are you to purchase this product if it were available today?”
Format: 5-point Likert. Survey-friendly. Industry benchmark: A top-2-box score (definitely/probably would buy) of 60 percent or higher is generally considered a positive signal for consumer products.
Q7. “What would this product need to include for you to consider buying it?”
Format: Open-ended. Best as a follow-up to Q6 in both the interview and the survey.
Q8. “At what price point would this feel like good value to you?”
Format: Open-ended or Van Westendorp price sensitivity scale. Red flag: If the expected price is significantly below your target margin, you have a positioning or value communication issue.
Comparative Questions
Q9. “How does this compare to what you currently use to solve this problem?”
Format: Open-ended. Best in the interview. Red flag: “It is basically the same” means your differentiation is invisible.
Q10. “What would make you choose this over your current solution?”
Format: Open-ended. Works in both. Best for identifying your actual competitive edge.
Q11. “What features do you find most valuable in this concept?”
Format: Ranking or open-ended. This directly shapes your product development roadmap.
Open-Ended and Qualitative Questions
Q12. “What are the three main reasons you would or would not use this product?”
Format: Open-ended. Works in both. This is one of the highest-signal questions you can ask.
Q13. “What improvements would make this more appealing to you?”
Format: Open-ended. Always include this. It turns a concept test into a roadmap.
Q14. “How likely are you to recommend this product to someone you know? (NPS scale)”
Format: 10-point NPS scale. Survey-friendly. NPS of 30 or above at the concept stage is considered strong for most B2C categories.
B2B-Specific Questions
If you are testing a concept for a business audience, these three questions are non-negotiable.
Q15. “Who else at your company would need to approve the purchase of a product like this?”
Format: Open-ended. Interview-friendly. Red flag: If multiple stakeholders are involved and your concept does not address their concerns, your sales cycle will be brutal.
Q16. “How would you justify the cost of this product to your finance team?”
Format: Open-ended. Best in the interview. This tells you exactly what ROI messaging to build.
How to Analyze Concept Testing Results
Collecting answers is the easy part. Knowing what those answers mean is where most teams stumble.
Quantitative Scoring
For purchase intent, use the top-2-box method. Count only respondents who answered “definitely would buy” or “probably would buy” as your positive signal. Anything below 60 percent for a consumer product warrants a rethink before moving to development.
For NPS, an NPS of 30 or above at the concept stage is a strong indicator. Below 10 is a clear warning sign. Industry NPS benchmarks vary, but these thresholds hold reasonably well across B2C product categories.
Qualitative Coding
For open-ended responses, do not read every answer in isolation. Group similar responses into themes. Look for the clusters that appear in 20 percent or more of responses. Those are your signals. The lonely one-off answers are just noise.
Red-Flag Patterns to Watch
- High interest, low purchase intent: This is the novelty trap. People find it fascinating but would never actually pay for it. Think Juicero.
- Widespread “I already have a solution”: This is a positioning failure, not a product failure. You need better differentiation messaging, not a different product.
- Low comprehension scores: If people cannot explain what the product does, the concept needs to be rewritten before any other data is meaningful.
- High satisfaction with current alternatives: Your concept needs a significantly stronger value proposition to move people away from habit.
When to Iterate vs When to Kill the Concept
Iterate when the core idea scores well but specific features, pricing or messaging are the problem. Kill the concept when the fundamental value proposition scores poorly across multiple respondent segments, comprehension is consistently low even after revisions or purchase intent top-2-box is below 30 percent after two rounds of testing.
Conclusion
Asking the right concept testing questions in the right order is not a nice-to-have. It is the difference between building something the market actually wants and spending six months on a product that ends up in a museum of failed ideas. Start with comprehension, move to value and only ask about purchase intent once people actually understand what they are buying. Map your questions to the stage of development you are in. And never skip the open-ended questions. The most useful insight you will ever get from a concept test will be in someone’s own words, not in a 1-to-5 scale. Articos helps teams build smarter research frameworks so that by the time you are committing budget to development, you already know it is going to land.
FAQs
A good concept testing question is clear, unbiased and actionable. It avoids leading language and tests one thing at a time. The answer should tell you whether to move forward, iterate or stop.
Between 8 and 15 questions is the sweet spot. At the discovery stage, stay under 10. At late validation, you can go up to 15. More than that and completion rates drop significantly.
Yes. Concept testing evaluates an idea before it is built. Usability testing evaluates a product or prototype that already exists. They answer different questions at different stages of development.
No. B2B concept testing needs questions about buying committees, ROI, and integration. B2C testing focuses more on personal relevance and lifestyle fit. Industry also matters; for example, healthcare questions look very different from fintech questions.
It reduces the risk of building something nobody wants. Companies that run structured testing protocols report failure rate reductions of 30 to 50 percent. It also gives you the evidence you need to get stakeholder buy-in before committing major resources.
