Don’t let average scores hide a fatal flaw

An idea can look great on paper and still have one weakness that makes it unworkable. Assess demand, feasibility and viability separately, and identify which uncertainties or risks could stop you.

Build & Validate
ExperimentationEvidence-led building

By the EWOR Team3 min read

Imagine comparing three startup ideas. One solves a problem customers care deeply about, but the technology is unproven. Another would be easy to build, but nobody has shown they would pay for it. The third looks commercially attractive, but the costs may make it impossible to deliver profitably.

Give each idea an average score and they might look equally promising. Yet each faces a very different challenge, and one could be far riskier than the numbers suggest.

Compare concepts side by side

Calling an idea “promising” does not tell you much about why it might work. Have customers repeatedly described the problem? Has an expert confirmed that the technical approach is plausible? Have you tested whether customers will pay enough to support the business?

These are different kinds of evidence, and none automatically proves the others. Customer interviews can establish that a problem exists without proving demand for your particular solution. An expert opinion can guide technical research without replacing a working test. An attractive price means little if the costs of delivering the product leave no viable margin.

Be clear about whether you have an unanswered question or evidence of a genuine weakness. An untested delivery model needs further investigation. Known costs that exceed what customers are likely to pay may call for a different approach. Giving either a numerical score will not resolve the uncertainty.

Show the evidence behind each assessment

“Promising” could mean customers have repeatedly described the problem, an expert believes the approach is technically possible or a founder likes the idea. These are very different foundations.

For each assessment, record the evidence behind it and what remains unknown. Discovery that confirms a problem exists does not establish demand for a specific solution. An expert's opinion on feasibility can guide further technical investigation, but it cannot replace a working test. An attractive pricing assumption does not prove the business can achieve healthy margins.

Distinguish an unknown from a demonstrated weakness. An untested delivery model needs further investigation, while known delivery costs that exceed a credible price range may require a redesign. Neither becomes clearer by assigning an artificially precise score.

Do not average away a blocker

A serious weakness in one area can outweigh strengths everywhere else. Before ruling out a concept, though, ask whether the problem lies in the underlying idea or simply in how you plan to deliver it.

Perhaps an expensive technical solution could be replaced with a simpler mechanism. Perhaps a different delivery model would make the economics work. If you change the underlying approach, treat it as a revised concept and assess it on its own merits rather than quietly changing the original assumptions.

Context matters, too. A business model that does not suit a venture-funded startup may still work for a business with different growth and financing goals. Judge each concept against the requirements of the venture you actually want to build.

Decide what to learn next

You do not need to settle every uncertainty before choosing what to investigate next. Ask which unanswered question could most change your preference between the concepts.

If one idea depends on unfamiliar technology, a focused expert assessment might help. If another relies on customers adopting a new workflow, you may need to investigate how they work today and whether they would change. Choose the next step based on what could genuinely change your decision.

Record why one concept deserves further attention, what evidence would change your mind and which alternatives remain worth considering. Then test the riskiest assumption behind your preferred option.

The goal is not to prove an idea with a scoring exercise. It is to avoid overlooking the weakness that could undo it, and to spend your next round of research where it matters most.

Get the next signal before everyone else.

Frameworks and contrarian theses for outlier founders, straight to your inbox. No fluff, no recycled advice, weekly.

The EWOR Signals newsletter