Put your ideas in front of the audience before you commit engineering, design or campaign budget. Find out which ones land, which fall flat, and why.
Concept testing answers one question: what do people think of this? Use it when you have a real idea, or several, and need to know which to back.
Accurate enough for the early decision about which concepts to back, not for the final launch gate. It shows which concepts synthetic participants understand, want and would act on, and why. What you get is a direction, not a percentage with a confidence interval, so screen the full set here and spend panel budget on the concepts that survive. For where that line falls, read when concept testing with synthetic users works.
Run concept testing when you have a candidate solution and a decision coming: a stage gate, a roadmap slot, a campaign to greenlight, or a shortlist that has to get shorter.
It guards against a familiar failure. The idea that won the loudest internal vote ships, and it turns out no segment wanted it.
you haven’t confirmed the problem the idea solves (run problem validation first), or the idea is settled and the open question is what to call it or how to describe it. That’s one of the three follow-on tests further down this page.
The format follows the number of concepts you bring.
The problem the idea solves is re-established first, so reactions are anchored to something participants care about.
The concept is introduced. After a first, unguided reaction, the interviewer checks the participant understood it before probing further.
Fit, the use they’d put it to, and their objections, probed however positive the conversation has been.
What would make them act on it, and for business buyers, how it compares with what they use today.
Head-to-head and card-sort interviews check comprehension and objections for every concept they go deep on.
Every concept gets one of five verdicts, in a decision table ordered best to worst that you can take straight into a roadmap conversation.
Alerts at the top when participants misread a concept, when they understood a concept but its name promised something the description didn’t deliver (the fix is to rename it, not to rewrite the description), or when a different group showed more pull than the audience you tested. How hard each objection is to answer, not only how often it came up. Needs none of your concepts covered, which is where the next idea usually comes from.
Head-to-head tests add what would have to be true for people to adopt, and for business buyers, who else has to sign off.
Everything above describes a standard test, where people meet whole concepts and react to them. The other three test options for one concept you already have. Participants meet the concept first, then see each option in that context, so nobody judges a name or a headline in a vacuum.
Show 1 to 12 whole concepts. People meet each one and react to it, and every concept gets a verdict from pursue to drop.
People meet your concept first, then react to 2 to 5 candidate names: does each one fit, and does it connect to what the product does?
2 to 5 benefit statements for one concept. Find the one people find most compelling, and the one they don't believe.
2 to 5 headlines or messages for one concept. Find the clearest one, and the one that makes people want to learn more.
Run them in the order the decision runs. Settle whether the concept lands with a standard test, then sharpen how you talk about it. Naming a concept nobody wants is effort spent on the wrong question.
Test ideas against a problem you already know is real.
People price something they already understand far better than something described to them moments earlier. Test the concept, then test the price.
Screen the full set of ideas synthetically, and spend panel budget on the ones that survive.
More detail gets sharper reactions. A one-sentence concept gets reactions to that sentence; a concept with its positioning, key features, audience and price gets reactions to each of those. The minimum is enough that a reader can say what it does, who it's for and why they'd use it. Candor tests text, so describe anything visual in words rather than relying on a mockup.
Yes, up to 12. The format follows the count. One concept gets a deep single-concept interview. 2 to 4 run head to head, with a forced choice at the end. 5 or more run as a card sort, where participants sort and rank the concepts and the interviewer goes deep on the top of the pile. The report orders every concept best to worst with its verdict, and head-to-head tests add what would have to be true for people to adopt.
A standard test asks whether an idea lands at all. A value proposition test starts from one concept you've already settled on and asks which way of stating its benefit people find most compelling and believable. Name and messaging tests work the same way: each tests options for a single concept, shown after participants have met the concept. Run the standard test first, because refining the words for an idea nobody wants doesn't help.
Yes, and the interview changes with the audience. For business buyers it probes the friction inside their organisation and, in head-to-head tests, who else has to approve. For consumers it probes how the idea fits the kind of person they are and how it makes them feel. The participants are built differently too: Candor models business buyers and consumers as separate domains, with different traits, biases and decision patterns. You choose the audience type when you describe who you want to understand.
It means participants didn't understand the concept, so any reaction to it would be a reaction to a misunderstanding. The report flags that instead of pretending to judge the idea, and shows where the description tripped people up. Fix the description and run the test again. It's the one verdict that says nothing about whether the idea is good, which is why it's kept separate from Drop.
Start with a free project, or walk through Candor with us first.