Test a product concept with real customers before you build it
Most product ideas are judged in a meeting room by people who will never buy them. By the time real customers see the product, the design is locked and changing it is expensive.
No credit card50 free respondentsGDPR & CCPA compliant
Concept testing shows the idea to the people you want to sell to, early, and measures whether they understand it, believe it and would buy it. It is the cheapest moment to kill a weak idea or fix a strong one.
SmartInterview combines quantitative surveys with AI-powered qualitative follow-up questions: you get purchase intent for each concept, and the reason behind every score.
Why concept testing matters
A concept test replaces opinion with evidence at the point where it costs least to change course. It tells you which concept to take forward and what to change in it.
✓Compare several concepts on the same metrics instead of debating them internally
✓Catch confusing wording or an unbelievable claim before launch
✓Know which customer segment is most interested, not just the average
✓Hear the exact objection that would stop someone from buying
How it is done today, and with SmartInterview
TodayWith SmartInterview
×Concepts judged in internal meetings by people who will never buy✓Target buyers rate each concept, and the AI asks why they hesitated
×Focus groups of eight where one loud voice steers the room✓Hundreds of individual answers, each probed by AI follow-up questions, no group bias
×Open comments pile up in a spreadsheet nobody has time to read✓Open answers coded into themes automatically, each theme linked to its verbatims
×Separate agencies per language, results arrive after the deadline✓Respondents answer in their own language, by voice or text, results within days
What you can learn from a concept test
Which concept wins
Appeal, purchase intent, uniqueness and believability for each concept, shown on its own (monadic) or in sequence, so the comparison is fair.
Why it wins
After each rating, the AI asks what exactly made the concept appealing or not. The answers are coded into themes, so the reasons are countable, not anecdotes.
What to change before launch
The barriers that recur, the wording people misunderstood, and the price at which interest drops, per segment.
How a concept test works
1
Define your research objective
Name the decision: which concept to develop, which variant to keep, or whether to launch at all.
2
Create the survey
Load your concepts as text, images or mock-ups. The questionnaire is drafted from your goal with standard concept metrics and a quality score.
3
Recruit respondents
Invite your customers or reach your target profile through a panel partner. Quotas keep each segment large enough to compare.
4
Analyze quantitative and qualitative insights
Read the ranking of concepts, the themes behind each score and the verbatims that show them, segment by segment.
Example studies
Customer story
Pharmalp: from pre-concepts to validated health products
Pharmalp, a Swiss natural-health company, had several product pre-concepts and needed to know which ones to launch and at what price.
Research design
Concept test with structured appeal and purchase-intent questions, a pricing module, and AI follow-up questions on every hesitation.
Respondents
Consumers in French-speaking Switzerland matching the target profile.
Business decision
Pharmalp selected the concepts to launch and validated their pricing and positioning.
Key insights
A clear winner emerged among the concepts
The acceptable price range was validated
The objections behind lower-scoring concepts were identified
Concept testing topics teams search for
Monadic vs sequential monadic testing
+
In a monadic test each respondent sees one concept, which gives clean scores but needs a larger sample. Sequential monadic shows several concepts one after another, cutting sample size at the cost of some order effects.
Purchase intent scale
+
The standard question uses a five-point scale from definitely would buy to definitely would not buy. Stated intent overestimates real purchase, so compare concepts against each other or a benchmark rather than reading the raw percentage as a sales forecast.
What is a top-two-box score
+
Top-two-box is the share of respondents choosing the two most positive points on a scale, such as definitely and probably would buy. It turns a five-point distribution into one number that is easy to compare across concepts and segments.
How to write a concept statement
+
A good concept statement names the consumer need, the product, the key benefit and a reason to believe, in under 100 words. Keep the format identical across concepts so the test measures the idea, not the quality of the writing.
Concept screening vs concept optimisation
+
Screening tests many rough ideas quickly on a few key metrics to drop the weak ones. Optimisation takes the shortlisted concept and tests variants of its features, price or wording to refine it before launch.
How to avoid order bias in concept tests
+
Rotate the order in which concepts appear, so no idea benefits from always being seen first. Asking the open question right after each rating, as SmartInterview does with its AI follow-up, also keeps each reason tied to the right concept.
Concept testing with mock-ups or images
+
Rough sketches and simple renders are usually enough at concept stage, as long as every concept has the same level of finish. A polished visual shown next to a text-only concept will win on presentation, not on the idea.
Qualitative vs quantitative concept testing
+
Qualitative testing explores why people react as they do, usually in small groups or interviews, while quantitative testing measures appeal on a larger sample. SmartInterview combines both in one survey: scores from every respondent, and AI follow-up questions that collect the reasons.
Pricing
Choose your respondent volume
Pro and Ultimate prices update instantly as you change the number of respondents per month.
Whether you collect one respondent this month or ten next quarter, you only pay for what you use. Your respondent credits are valid for 3 years.
Frequently asked questions
Concept testing is a research method that shows a product or service idea to target customers before it is built, to measure how appealing, clear, believable and purchasable it is, and to understand why.
Describe the concept in a short text or image, show it to people who match your target customer, measure appeal and purchase intent, then ask open follow-up questions to understand what drives each answer.
Around 100 to 150 respondents per concept or per segment gives a reliable comparison. Fewer, around 20 to 30, are enough to understand the main reasons behind the scores.
Overall appeal, purchase intent, uniqueness, believability, relevance and price perception, followed by open questions such as what they liked most, what would stop them from buying and what they would change.
As early as a concept can be described clearly, before design and production are locked, and again after each major revision.
Concept testing evaluates an idea before the product exists. Product testing evaluates the real product once people can use or taste it.
Yes. AI can draft the questionnaire, ask follow-up questions when an answer is vague, and code thousands of open answers into themes. The decision on which concept to launch stays with your team.
On SmartInterview the first 50 respondents are free. Paid plans start at CHF 45 per month for 150 respondents per month. Recruiting through a panel partner is priced separately.
A short test can be live the same day. With your own customers, most teams read the first results within a few days.
Yes. Respondents answer in French, German, Italian or English, and the analysis is consolidated across languages.
