Concept Testing: Should Each Person See One Idea or Both?

Someone who sees your refill pouch after your reusable bottle has already learned something about your product range. Someone who sees only the pouch has to make sense of it alone. Give them the same survey question and their answers can still mean different things.

I’d show each person one concept when the decision depends on how an idea stands on its own, and both concepts when the decision depends on how people compare them. A smaller recruitment budget can affect that choice, but it shouldn’t quietly change the question your study answers.

What is monadic concept testing?

Concept testing means showing your target audience an idea before launch, usually through a description, image, mockup or video, then asking what they understand, how they react and what they might do.

When each participant sees just one concept, that’s monadic testing. To test two concepts, you recruit two comparable groups. One sees A; the other sees B. Both groups answer the same questions about the concept they saw.

When each participant sees multiple concepts one after another and rates each separately, that’s sequential monadic testing. For two concepts, some participants should see A then B, and others B then A. You can add a direct preference question after both sets of ratings.

Showing both concepts together and asking people to pick one is a direct comparison. It can be useful, but it doesn’t give you a separate first reaction to each idea.

These definitions follow Sogolytics’ September 17, 2026 concept-testing guide and SurveyNinja’s September 30 guide to monadic and sequential monadic testing. Both are vendor guidance, not comparative validation studies. They help name the designs; they don’t establish that one always produces better decisions.

Which design fits your decision?

Before writing questions, finish this sentence: “After this study, I need to decide whether to…”

If the answer is “put concept A or B in front of customers as a standalone offer,” I’d start with one concept per person. You need to know whether customers understand and want each offer without help from the other.

If the answer is “choose which of these two benefits to emphasize” or “understand which trade-off customers prefer,” showing both may be more useful. People can explain what they gain or give up by choosing one.

Consider the eventual buying situation, too. If shoppers will see both options on the same product page, comparison is part of the experience you’re trying to understand. If they’ll arrive on a landing page featuring one offer, a standalone reaction is closer to that encounter. Neither survey recreates an actual purchase, but you can avoid adding a comparison customers won’t have.

I’m tempted by showing both because every participant can comment on every idea. That feels efficient. It also makes it easy to mistake “preferred over this other option” for “appealing enough to buy.” Two weak concepts can still produce a winner.

How exposure changes an answer

Think about tasting two coffees. If you taste only one, you judge it against your habits and memories. If you taste a very bitter coffee first, the next cup may seem mild by comparison. The second cup hasn’t changed; your reference point has.

Concepts work in a similar way. The first idea can teach participants terminology, draw attention to a benefit or create an expectation about price. The next idea gets judged in that context.

With one concept per person, you avoid exposure to the other tested concept. You still have differences between people. One group might contain more frequent category buyers or more loyal customers of your brand.

Random assignment helps reduce systematic differences between groups. Use the same recruitment criteria and source for both, then check whether relevant characteristics ended up reasonably comparable. Random assignment doesn’t guarantee perfectly matched small groups.

With both concepts per person, you can compare reactions within the same individual. Someone who uses the top end of every rating scale does so for both ideas, which can make their relative response easier to interpret. But you introduce order effects, learning and fatigue.

Randomly assign participants to A-then-B or B-then-A, aiming for roughly equal groups. This prevents one concept from always benefiting or suffering from the same position. It doesn’t erase exposure effects. Keep the order attached to each response so you can inspect it later.

Comparable concept descriptions

Your study design won’t rescue an uneven pair of stimuli. If A gets a polished photo and a vivid benefit while B gets a rough sketch and a vague sentence, you’re testing those presentation differences too.

Give both concepts the same structure, similar detail and equivalent visual treatment. Include the information customers need to judge them. If price is part of the intended offer, show it consistently. If you leave price out, interpret purchase-intent answers as reactions to an unpriced idea.

Keep genuine differences visible. Making descriptions comparable doesn’t mean removing the feature you need to test.

Decide whether branding belongs in the test. A branded concept answers a question about that offer from your brand. An unbranded concept leaves brand associations out. Use the same choice for both unless branding itself is the variable.

Pilot the material with a few people before recruitment. Ask them to explain the offer back to you. If they misunderstand a shared instruction, fix it in both versions before collecting the main responses.

An illustrative pair of concepts

The following concepts and study designs are illustrative. They describe an imagined DTC cleaning brand, not a customer study or observed results.

The brand wants to choose a refill format for a surface cleaner. Both descriptions use the same photo style, scent, cleaning claims and price. The refill format and preparation steps differ:

• Concept A, ready-to-use pouch: “A refill pouch containing 500 ml of ready-to-use surface cleaner. Pour it into your existing spray bottle. One pouch fills the bottle once.”
• Concept B, concentrate sachet: “A concentrate sachet that makes 500 ml of surface cleaner. Pour it into your existing spray bottle, add water to the fill line and mix. One sachet fills the bottle once.”

Those quantities define the imagined offers. They aren’t performance claims about a real product.

The brand’s decision could be framed two ways. “Which format makes sense and appeals on its own?” calls for the first outline below. “How do customers weigh ready-to-use convenience against a concentrate?” calls for the second.

Outline 1: One concept per person

Decision: Choose which refill format to develop as a standalone offer.

Audience: People who buy household surface cleaner and use a refillable spray bottle. Apply the same eligibility rules to both groups. Avoid recruiting one concept’s group from loyal customers and the other from an unfamiliar panel.

Assignment: Randomly assign each eligible participant to A or B. Each sees only the assigned concept.

Questions: Ask these in the same order, with identical wording and response options:

1. “In your own words, what would you receive, and how would you use it?”
2. “How appealing is this refill option to you?” Use a fully labelled five-point scale from “Not at all appealing” to “Extremely appealing.”
3. “At the price shown, how likely would you be to buy this refill the next time you need surface cleaner?” Use a fully labelled five-point scale from “Very unlikely” to “Very likely.”
4. “What most influenced your answers?”

Keep the concept available while participants answer. The first question checks what they take from the description; it needn’t become a memory test.

Reading the results: Compare comprehension, appeal and stated purchase intent across the two groups. Read explanations alongside the ratings. A concept that scores well because people misunderstand what arrives in the package needs revision before you treat its score as support.

Choose your decision rule before seeing results. For this brand, I’d require people to understand the preparation steps before letting purchase intent drive the choice. If reactions are close or uncertain, revise the unclear parts rather than forcing a winner.

This design can support a standalone comparison. It won’t tell you which specific trade-off the same person would make after seeing both formats.

Outline 2: Both concepts per person

Decision: Understand which format people prefer after considering both, and why.

Audience: Use the same eligibility criteria as the first outline.

Assignment: Randomly assign eligible participants to A-then-B or B-then-A. Show one concept at a time and collect its ratings before revealing the next.

Questions: Repeat the same four-question block from Outline 1 after each concept. Then show both again and ask:

9. “Which would you choose at the prices shown?” Offer A, B, “Either, no preference” and “Neither.”
10. “What most influenced that choice?”

Don’t ask participants to compare the concepts before they have rated both. Avoid “Which is better?” because it leaves people to invent the criterion and nudges them toward choosing a winner.

Reading the results: Look at each person’s ratings together, then their final choice and explanation. Separate people who prefer A from people who reject both. A forced A-or-B answer would hide that difference.

Inspect results by presentation order. If B looks stronger mainly when it follows A, report that pattern rather than folding everything into a single confident verdict. Small order groups may only support a tentative interpretation.

You can also examine first-position responses, where participants haven’t yet seen the alternative. Those subsets contain fewer responses, and the survey introduction may already have told people to expect a comparison. Keep those limits visible.

The sample and analysis trade-offs

Splitting a fixed sample means fewer people rate each concept. In an illustrative study with 80 participants, an even monadic split gives you 40 ratings for A and 40 for B. Showing both gives you up to 80 ratings for each, assuming everyone completes both sections.

That arithmetic isn’t a minimum-sample recommendation. The useful size depends on the decision, the differences you need to detect and the uncertainty you can tolerate.

Nor does the second design create 160 independent participants. Each person’s two ratings are connected. If you use statistical tests, use methods that account for those paired responses; for separate groups, use methods suited to independent samples.

Showing both also asks more of each participant. Budget for the longer questionnaire and check completion patterns. A longer study may affect recruitment costs, so compare actual quotes rather than assuming two ratings per person cost the same as one.

Whatever the design, stated purchase intent remains a survey answer. It can guide the next test. It cannot establish sales or prove demand.

How Probe fits the study

In Probe, you can design a survey or interview study and send it to your own audience, or have us recruit participants from our recruitment partner’s vetted panel. Own-audience studies are free up to 10 questions and 25 responses. For recruited studies, you see one all-in price per study before launch. Results come back as charts, themes and quotes.

I’d use a survey for the structured outlines above. Interviews can help afterward if you need to understand an unexpected interpretation or a trade-off people struggle to explain in writing.

Confirm the assignment and presentation-order setup before launch. The design requires random assignment, but this guide isn’t a claim that Probe automatically implements those experimental controls. For your own audience, separately prepared versions with an external random allocation process are one possible route. Keep each person in their assigned version and record which version they received. For recruited work, confirm that the required allocation can be supported before paying.

The both-concepts outline contains ten questions before adding screening or background questions. Don’t assume the complete study fits the free limits once those are included. If your implementation requires separate recruited studies, each has its own quoted price.

Keep the decision written down alongside the design. How to Write a One-Page Research Decision Brief gives you a place to state what evidence would change your choice.

Before building the survey, write two possible findings: “People understood and wanted A when they saw it alone” and “People preferred A after seeing B.” Choose the finding you actually need. Then make the exposure sequence match it.