How to evaluate a product idea before you build it
Published · By Daily Product Idea · Our methodology
The short answer
Evaluate a product idea by scoring the evidence, not your enthusiasm. Rate problem evidence, customer clarity, market potential, differentiation, buildability and customer access; write down why each score is what it is; list the assumptions the idea depends on that no source has established; and design the cheapest test for the riskiest one. A low score honestly reached is more useful than a high one you cannot explain.
Score the evidence, one dimension at a time
An idea you like will always look good as a whole. Break it into parts and judge each against what you can actually show. Daily Product Idea scores every researched opportunity on six dimensions, each from 0 to 100, weighted into an overall opportunity score:
| Dimension | Weight | The question it answers |
|---|---|---|
| Problem evidence | 25% | How many independent, specific, recent sources show the problem, and how many show a consequence or workaround? |
| Customer clarity | 15% | Can you name the customer precisely enough to find them, and do the sources agree on who they are? |
| Market potential | 15% | Is there evidence of spending, paid manual work or search demand, or only of complaint? |
| Differentiation | 15% | What already exists, including free tools and platform features, and what would this do that they do not? |
| Buildability | 15% | Can a small team build the smallest useful version, and what does it depend on that you do not control? |
| Customer access | 15% | Where exactly would you find the first twenty customers, and at what cost? |
Problem evidence carries the most weight because everything else depends on it: a well-differentiated, easy-to-build product for a problem nobody has is still worth nothing.
Write down the reason for every score
A score without a reason cannot be checked, and a score you cannot check tends to drift upwards. For each dimension, write one or two sentences that point at specific evidence, and say explicitly where evidence is thin or missing. Then score that dimension low rather than rounding up out of optimism.
Flight Deck shows why. Its problem evidence scores 70: three separate issue-tracker threads describe the same difficulty. Its differentiation scores 20, because the research found several free tools covering most of the proposed product, and its market potential scores 25, because no sign of anyone paying for such a tool was found. The overall score of 49 is low, and the reasons tell a builder exactly where the opening, if there is one, would have to be.
Separate findings from assumptions
Every idea rests on beliefs that no source has established. Write them down as a separate list, in plain words, and never let one slip into the evidence. For Trade Reality Dossiers, a paid report built from interviews with people running a trade, the research found evidence that first-time business buyers face real risks before committing, and that people pay for interview-based business research. It found nothing at all on whether small operators would share their real margins and supplier names with a publication. That is an assumption, and buildability is scored 30 because of it.
Evidence or hypothesis?
Design the cheapest test for the riskiest assumption
The point of evaluating an idea is to decide what to do next, and the answer is almost never "build the whole thing". Rank the assumptions by how badly the idea fails if each is wrong, and design the cheapest test for the top one. Good first tests are small, fast and capable of saying no:
- Deal Floor: before building beyond a static page, ask eight to ten owners to talk through their last three quotes and see whether they can produce an overhead figure and close rate at all. If they cannot, the calculation fails at its first step.
- Trade Reality Dossiers: publish a free teardown and see whether a waitlist converts at the intended price, before doing any interviews.
- Flight Deck: build only the part that tells "needs you" from "finished", and stop if it cannot beat the noisy signal described in the issues.
What an overall score means, and what it does not
We read overall scores in bands:
- 80–100: Strong opportunity, still requires testing
- 65–79: Promising
- 50–64: Mixed
- 35–49: Needs a sharper angle
- 0–34: Not currently recommended
A score is a summary of the evidence at one point in time, under one version of the method. It is not a prediction of success, and a high score still means "worth testing", not "validated". Only direct evidence from customers, such as interviews, sign-ups, pre-orders or observed use, can justify calling an idea validated.
A short checklist
- Score each of the six dimensions separately, from the evidence you can point to.
- Write a reason for every score, naming the evidence and the gaps.
- List what counts against the idea next to what supports it.
- List the assumptions separately from the findings.
- Rank the assumptions by how badly the idea fails if each is wrong.
- Design the cheapest test that could prove the riskiest one wrong, and run it first.
Every idea on Daily Product Idea is published with its six scores, its sources, its risks and the tests to run first. See how ideas are researched and assessed or browse the ideas.