Most product teams have more ideas than development budget. In a gated new product development process, concept screening sits at the front end: the first gate, where a long list of early ideas is narrowed to the few that merit more work. It answers a narrow, practical question: if this product existed at this price, would the people who buy in the category want it, and would it bring in new business or just move existing customers from one of your products to another?
Where it fits in development
Concept scores are rarely used on their own. Most teams set them beside market potential, fit with what the organization can make and sell, and fit with strategy. The ideas that hold up on all of those become the product development roadmap. From there, each surviving idea passes through more gates, with richer stimuli at each one:
- Written concept boards (the test on this page)
- Concepts with photography or illustration
- 3D renderings
- Hand-feel prototypes, for look, size and feel
- Functional prototypes
- Home or shop tested prototypes, used in real conditions before launch
Each gate costs more than the one before it. That is why the first one matters: it keeps weak ideas from using up rendering, tooling and prototype budget.
What a concept is
At this first gate, a concept is a short written board: a name, a one-line idea, three benefits and a price. No photography and no packaging design yet, so the test measures the idea rather than the styling. Every board is written to the same length and reading level, and your current best seller is written up the same way so it can serve as the benchmark.
How we run it
We recruit people who have bought in the category recently and split them into matched groups of about 300, one group per concept. Each person sees one board only. This is called a monadic design, and it matters: when people see several ideas side by side they start comparing them to each other instead of to the shelf they actually shop. Each board is rated on purchase intent, then on the diagnostics that explain it: uniqueness, believability, relevance, value for money and likely frequency. Two open questions ask what people like and dislike, and likely buyers tell us what the new product would replace.
What you get
A readout that leads with a recommendation, not a pile of tables. Every score is shown against the benchmark product and a norm band from earlier tests in the category, with significance letters so you can see which gaps are real and which are noise. Open-ended answers are coded into themes with verbatims beside them. Purchase intent is broken out by the segments that matter to you, and a source of volume view shows how much of each idea's appeal is new to your brand. The last page lists what to test next.
Where it goes wrong
The most common mistake is reading purchase intent alone. A unique idea nobody believes, or a cheap idea that only cannibalizes your own range, can post a respectable top line and still be the wrong one to build. The second mistake is uneven boards: one concept gets a better name or a lower price and wins for that reason. The third is treating a concept test as a forecast. It picks the idea. It does not prove the product, which still has to pass the later gates and taste, work or perform the way the board promised.
The sample below is a full readout from a small screen: a fictional frozen meal brand testing four early ideas against its current best seller. A real front end screen often covers many more. Click the tabs to move from the summary to the boards, the scorecard, the open ends, the segments and the source of volume. Every brand and number is invented.
Illustrative only. All brands, products, people and figures are fictional. Open the sample full screen.