Scoring
Nine dimensions, and what evidence each one wants
A score is only useful if you can see what produced it. This is the whole rubric: what each dimension asks, what kind of evidence moves it, and the failure that most often costs points.
Why nine and not one
A single number hides the disagreement that matters. An idea can have obvious demand and no way to reach the people who want it; another can have a clean business model and three funded incumbents already serving the use case. Both average to something unremarkable, and the average is the least useful thing you could be told.
Nine dimensions keep the disagreement visible. The verdict comes from the shape of the profile, not from the sum, which is why a report shows you the bars before it shows you the number.
What each dimension measures
Every dimension is scored against evidence that was looked up, not against how the idea reads. Where the evidence does not exist, the dimension scores low and says so — an unknown is not a neutral.
| Dimension | What it asks |
|---|---|
| Market demand | Is there evidence that people already look for this, pay for something adjacent, or complain about its absence? Search volume, existing spend and community complaints all count; enthusiasm from the founder does not. |
| Problem intensity | How much does the problem hurt, and how often? A weekly irritation and an annual crisis score very differently, and both differ from something people merely agree is suboptimal. |
| Market opportunity | Size, growth and how much of it you can actually address. A large market you cannot reach scores below a small one you can. |
| Competitive advantage | What stops the third incumbent from doing this next quarter? Distribution, data, switching costs and regulation are defences; being first is not. |
| Business model clarity | Who pays, how much, and how often — and whether the unit economics survive the cost of acquiring that payer. |
| Effort-to-value ratio | What it costs to build the first useful version against what that version delivers. A twelve-month build to test a hypothesis scores badly however good the hypothesis is. |
| Distribution feasibility | Whether there is a channel you can afford. This is the dimension that most often ends an otherwise strong profile. |
| Strategic fit | Whether the person building it has the context, access or unfair advantage that this particular idea needs. |
| Timing | What changed recently that makes now different from three years ago. A regulation, a price collapse, a platform opening. Without an answer, this is the dimension that quietly explains why nobody has done it. |
What makes a score fall
The commonest reason a dimension scores low is not that the answer is bad, but that no evidence was found either way. That is deliberate. A score built on the absence of contradiction is a score that flatters every idea equally.
The second commonest is a mismatch between two dimensions that each look fine alone. Strong demand with no affordable channel, or a clean model with no defence — these show up as a profile with a hole in it, and the hole is the finding.
What the score is not
It is not a prediction, and it does not know your execution. Two people with the same idea and the same score have very different odds, and nothing here measures that difference.
It is a reading of the evidence available at the moment it ran. That is why every report cites what it found, and why the same idea can score differently three months later — which is the whole reason the Radar exists.
The first analysis is free and needs no card.