Concept Testing SOP

A concept testing SOP shows your solution ideas to real customers before you design or build anything, so you spend your prototype budget on the idea people understand and want. It takes one to two weeks. Each idea becomes a concept board: one page with a headline, the problem, how the solution works, a picture and a price. Ten minutes of a customer's honest confusion on a concept board saves weeks of building the wrong screens.

What is the concept testing SOP and when should you use it?

Concept testing comes from consumer goods research, where brands test product ideas on boards long before production. You measure four things: whether people understand the concept, whether they want it, whether it beats what they do today and whether the price feels acceptable. Two survey designs are standard. In a monadic test each respondent sees one concept, which gives clean scores but needs more people. In a sequential monadic test each respondent sees every concept in random order, which needs fewer people and makes comparison easier.

Purchase intent uses a five-point scale from "definitely would buy" to "definitely would not buy". Research on purchase intentions, including work by marketing professor Vicki Morwitz, shows stated intent runs higher than real buying. That is why this SOP judges concepts by the top box ("definitely would buy") and by how concepts compare with each other, and treats "probably" answers as noise.

Use this SOP:

• When you have two or more solution ideas for one validated problem and can't choose.
• After a problem-solution fit canvas, to test the solution block with more people.
• Before a prototype test, since prototypes cost more and measure usability once you know what people want.

Scope check: concept testing compares ideas in a controlled setting. A landing page test measures signups from live traffic, and solution value proposition testing compares messages for one chosen solution in real channels.

What do you need before you start?

• A validated problem and one customer segment.
• 2 to 4 solution ideas that solve the problem in different ways, with more than new wording.
• A design tool (Figma or Canva) for the boards; no design skill needed beyond clean layout.
• A survey tool (Typeform, Lyssna) and a way to reach 100+ people in your segment: your own list, communities or a panel such as Prolific.
• 8 people from the segment for short qualitative calls.

How do you run it, step by step?

1. Write each concept in the same format

Every board gets the same five parts: a headline of under ten words, the problem in one sentence of customer language, how it works in three short steps, one image (sketch, mockup or photo) and a price. Keep length and design quality equal so the idea wins on its own merits.

2. Run 8 qualitative calls first

Show each board for 30 seconds, hide it, and ask the person to explain it back in their own words. Then ask what they liked, what worried them and how it compares with what they do today. Fix any board that two or more people misread before the survey goes out.

3. Choose monadic or sequential monadic

With 300+ reachable respondents, run monadic: 100+ per concept. With fewer, run sequential monadic with 100+ total and randomise the order. Note the order effect: the first concept seen tends to score higher, which randomising cancels out.

4. Build the survey

Include a screener, the concept board, a comprehension check, appeal (1 to 5), purchase intent (five-point), "better or worse than what you do today" and a price question. End with one open question: "What would stop you from using this?"

5. Field the survey and watch quality

Collect responses over 5 to 7 days. Remove speeders (people who finish in a third of the median time) and anyone who fails the comprehension check badly, and note how many you removed.

6. Score and compare

For each concept compute comprehension rate, top-box and top-two-box purchase intent, share who say it beats today, and price acceptance. Read every answer to the "what would stop you" question and tag the objections.

7. Pick, merge or drop

Pick the concept that leads on top-box intent and comprehension. If two concepts are close, check whether their best parts merge into one. If none passes your preset thresholds, go back to the canvas before you prototype anything.

Template you can copy

CONCEPT BOARD (one page, same layout for every concept)
HEADLINE: [under 10 words, the outcome]
THE PROBLEM: [one sentence in customer language]
HOW IT WORKS: 1. ____  2. ____  3. ____
[ IMAGE: sketch / mockup / photo ]
PRICE: [€__ per month / per use]

SURVEY
S1–S2. Screener questions (end if not in segment)
Q1. In your own words, what does this offer do? (open text)
Q2. How appealing is this to you? 1–5
Q3. How likely would you be to buy it at the price shown?
    Definitely would / Probably would / Not sure / Probably not / Definitely not
Q4. Compared with how you handle this today, is it: Much better / Better /
    About the same / Worse
Q5. Is the price: Too high / A bit high / About right / Low
Q6. What would stop you from using this? (open text)

SCORING SHEET
Concept | n | Comprehension % | Appeal avg | Top-box % | Top-two-box % |
Better than today % | Price "about right" or lower % | Top 3 objections

Worked example: dog owners and vaccinations

This is an illustrative case. A founder has three ideas for busy dog owners who miss vaccination dates: A, a vet chat subscription at €9 a month; B, at-home vaccination visits at €49 a visit; C, a free pet health record app with paid reminders at €3 a month. Eight calls show that people misread C as a vet booking app, so she rewrites its headline.

She runs a sequential monadic survey with 150 dog owners from a panel, order randomised. Comprehension: A 88%, B 92%, C 79%. Top-box intent: A 14%, B 27%, C 6%. "Better than today": B 71%, A 48%, C 33%. Price: 64% call B's price "about right" or lower.

The top objection to B is "I want my own vet", raised by 31 people. She decides to test B as a service delivered with local vets instead of her own staff, and moves to a landing page test to see whether interest holds when people have to leave an email.

What should you measure, and when do you stop?

Stop once the winner holds its lead after you remove low-quality responses. If no concept passes, the problem may be right but the solutions miss a constraint: return to the problem-solution fit canvas and check the constraints block.

Which mistakes should you avoid?

• Uneven boards. A polished mockup next to a rough sketch tests design skill. Match the quality.
• Leaving out the price. Without a price every concept sounds great. Show the price you plan to charge.
• Counting "probably would" as demand. Stated intent overstates buying. Judge by top-box and relative ranking.
• Testing one concept alone. One concept has nothing to beat. Test at least two, or one concept against the customer's current method.
• Skipping the open question. The objections tell you what to fix before the landing page.

How can AI speed this up?

An AI assistant can turn rough ideas into boards in the same format, write five headline options per concept and check that all boards have equal length and reading level. After fielding, it can tag objections and compute the scoring table from your CSV; spot-check ten tagged answers. Image generators produce quick board visuals, and you should keep them all in one style. For board copy, try the free product description generator and the elevator pitch generator, then edit the output into customer words.

Where does this SOP fit in your validation plan?

Concept testing opens Solution Validation. Before it comes the problem-solution fit canvas, which closes problem work. After it, test live demand with a landing page test, tell the story of the winner with solution storyboarding, check its name with solution name testing and test usability with a prototype test.

FAQ

What is concept testing in product development?

Showing customers a short description of a product idea, before it exists, to measure whether they understand it, want it and would pay the stated price. It helps you choose which idea to prototype.

How many people do I need for a concept test?

About 8 for qualitative calls, then 100+ per concept in a monadic test or 100+ in total for a sequential monadic test.

What is the difference between monadic and sequential monadic testing?

In monadic testing each person sees one concept. In sequential monadic testing each person sees all concepts in random order. Monadic gives cleaner scores; sequential monadic needs fewer respondents.

What is a good purchase intent score?

There is no universal number, because stated intent always runs above real buying. Set a top-box bar before you start and focus on how far the winner leads the other concepts.