Concept testing examples: what to test after the first result
A modeled protein-drink study and four fictional examples show how audience fit, misunderstood terms and buying occasion change the next test.
A concept test should change what you develop, whom you test with or how you explain the offer. In a modeled protein-drink study, the buyer's existing routine points to a first paid-trial audience. The service, meal-planner, bottle and tea examples that follow are fictional; they show how different responses call for different next tests.
Published modeled case: choose a first trial audience
Protein habit changes the next test
Our published modeled protein study first compared three unofficial formats, then focused on a 330 ml clear citrus drink with 20 g protein, 80 mg caffeine and a $3.99 price. The follow-up modeled 2,000 U.S. adults and was not commissioned by Red Bull.
Existing protein users had modeled first/regular purchase rates of 31.1%/5.2%, versus 21.5%/2.9% without a regular protein habit. First purchase meant full-price buying within eight weeks; regular buying meant at least four cans in weeks 5–8 after that purchase, across two shopping trips. Separate subgroup bases and uncertainty intervals were not supplied.
The next action is a paid product test starting with existing protein buyers and relevant caffeine routines. Taste, observed repeat and displaced purchases remain to measure. The lesson is to define the first audience and behavior, rather than treat broad concept interest as a launch forecast.
Service case: a fee misunderstanding comes before route planning
MendRoute is an appliance-collection service for households in its listed service area. It collects one eligible small appliance in a listed neighborhood, takes it to a repair partner for assessment and returns it. Collection and return cost $12. Repair is quoted separately and may be declined; neither a repair outcome nor a completion date is guaranteed.
The $12 service fee needs an explicit boundary
Answer to investigate
“Twelve dollars fixes it and brings it back?”
Candidate card revision
$12 covers collection and return. Repair is assessed and quoted separately. You may decline that repair quote. Check the listed eligible appliances and service area before booking.
No repair outcome or completion date is guaranteed in this concept.
The audience is service-area households with a broken eligible appliance in the past three months. Ask what the $12 covers, what would happen next, what they would do instead and what information is missing. Those questions distinguish price misunderstanding from a lack of collection need.
The answer “Twelve dollars fixes it and brings it back?” imagines an offer the business has not made. Another answer needs appliance eligibility; a third can drop the item off themselves and sees little value in collection. These are three different reasons to hesitate, not one generic lack of appeal.
The candidate card puts collection and return beside the separately quoted repair, preserves the right to decline and makes eligibility and area visible. Retest understanding before appeal; the effect on booking hasn't been measured. If people understand the offer but don't need collection, revisit who the service is for instead of rewriting the same fee explanation indefinitely.
Committing collection routes also needs operational evidence. Map feasible coverage and repair-partner responsibilities before making that commitment. Opinion research can refine the offer and audience; it cannot prove that the service can fulfill its promise. Revise the card first, then use the new answers to decide whether the target needs to change.
Message case: interest in the wrong product
Weektable must sound like a planning app
Weektable is the example meal planner for household shoppers. You need its headline to communicate planning and an ingredient list rather than prepared-food delivery.
Concept | Description and mechanism |
|---|---|
A | Description: “Dinner, without the daily debate.” Intended mechanism: A weekly meal plan |
B | Description: “Pick five dinners. Get one organized grocery list.” Intended mechanism: Planning plus an ingredient list |
Some answers in the example to “Dinner, without the daily debate” imagine prepared meals delivered to the door. The more explicit alternative says to choose dinners and receive an organized grocery list. No counts or statistically established winner are supplied.
The emotional benefit still needs to point to the app's actual job. Interest in meal delivery doesn't tell you whether someone wants a planning app. Keep the product and price fixed while testing a clearer explanation.
A candidate message is “Plan five dinners, get one grocery list, and spend less time deciding what to cook.” Retest unaided understanding before a relative headline choice. Ask what people expect to receive and do themselves. If they still imagine prepared food, revisit the presentation rather than declare the attractive phrase successful.
This is a messaging decision. A full concept test would also examine the complete planner offer, use occasion, price and alternatives. Choose the study scope around what is actually unresolved.
Two comparisons show why the base matters
Bottle performance and tea buying occasion
Desk Brew and Travel Brew are 500 ml reusable tea-bottle offers at $25 online. Desk Brew's card describes a wide opening for easy cleaning; Travel Brew's describes a leak-resistant lid for commuting. In the example's 100-response cells, positive relevance is 48 versus 62 and positive intent 39 versus 44 for Desk and Travel respectively: gaps of 14 and 5 percentage points. The larger relevance gap makes Travel worth investigating; test the actual lid before making a leakage claim.
The tea occasion cards describe the same lightly sweet sparkling tea, six 250 ml cans at $15 in grocery. Among 100 afternoon drinkers in the example, the afternoon direction wins 65–35. Among 100 dinner-beverage seekers, dinner wins 60–40. Equal blending gives afternoon (65+40)/200=52.5%, concealing the group fit.
The example's reasons point to a tea-break alternative for afternoon drinkers and a meal occasion without an afternoon break for dinner-beverage seekers. Choose a reachable audience before selecting its occasion message. These bottle and tea comparisons use separate audiences from the published protein study.
A relevance gap, a blended preference and a modeled repeat rate answer different questions. Use each to choose its own next test, rather than pool them into a generic success score.
Each next test should resolve the named gap
MendRoute needs the revised fee and eligibility card with the same comprehension question. Weektable needs a clearer mechanism with the same product. The protein case needs a paid product trial with repeat and substitutions defined. Each next study has an input and a question that follows from the previous evidence.
State the branch before collection. Clear understanding plus weak need calls for audience work. Continued misunderstanding calls for another presentation change. Favorable opinion plus unknown performance calls for a product test. Those branches make the examples useful across categories without transferring their prices or rates.
The recommendation needs a traceable response
Retain the card, audience, exposure and response source beside the interpretation. Name the change and the remaining evidence. This lets a reviewer assess whether the recommended next step addresses what people actually saw and said.
The useful result is a next step that matches the evidence: a product trial for the protein drink, a clearer fee for MendRoute or a clearer mechanism for Weektable. Keep modeled reactions, human opinions and observed purchases labeled, because each leaves a different question unresolved.
Further reading
For an offer that people misunderstand, discuss a Vensia concept study focused on the explanation and audience you need to test next.
