Why a confidence score at all?
The hard part of evaluating an idea is comparing it. A score, even an imperfect one, lets you scan dozens of ideas and triage which deserve real validation. The Vibe Ideas confidence score is a single 0–100 number weighted from five signals. We never claim it predicts success — we claim it ranks ideas consistently, which is what a busy founder actually needs at the start of an evaluation.
The five signals
- Market depth — is the TAM large enough that even a small share is a real business?
- Acquisition feasibility — can you reach the audience cheaply, or is CAC structurally high?
- Build feasibility — can a small team ship the MVP in weeks, not quarters?
- Monetisation clarity — is there a known willingness to pay, or do you have to invent demand?
- Defensibility — is there at least one moat hypothesis (data, network, distribution, brand)?
Each is scored 0–20. We deliberately omit "founder fit" and "timing" because both are subjective and undermine the score's usefulness as a comparison tool.
How the weights work
We weight all five signals equally because no single signal reliably outperforms the others across categories. A weak market with strong distribution can absolutely succeed (think single-creator SaaS), and a strong market with weak distribution can absolutely fail. Equal weighting protects you from over-fitting to the latest fashionable signal.
Reading the score
- 75–100 — strong on most axes; worth running a validation sprint.
- 55–74 — viable but with at least one structural weakness; investigate before committing.
- 40–54 — workable for a side project, risky as a full-time bet.
- 0–39 — usually a hobby or a feature, not a business.
A 60 is not "60% chance of success" — it is "this idea ranks in the upper-middle of generated ideas in the same category." Treat the score as a sorting tool, not a prediction.
What the score deliberately ignores
The score does not include founder fit, timing, geographic constraints, or capital availability. Those matter enormously, but they are personal, not properties of the idea. We keep them out so the same idea gets the same score regardless of who is reading it. You should layer your own personal-fit filter on top before deciding which idea to commit to.
How to use the score in practice
Generate 20 ideas, sort by score, kill the bottom 10 immediately, run a 5-minute gut check on the next 7, and reserve the top 3 for a real validation sprint. The score is a triage tool — it earns its keep by saving you days of evaluating ideas that would have been killed in the first conversation anyway.
Common misreadings
- "Score 90 means I should quit my job." It does not. It means an interview-driven validation sprint is justified.
- "Score 30 means the idea is bad." Not necessarily — it means structural weaknesses need a counter-argument.
- "Two 70s are equivalent." Not when they get there via different signal mixes. Read the breakdown, not just the total.
Key takeaways
- The confidence score is a 0–100 sort key built from five equally-weighted signals.
- It deliberately ignores founder fit and timing so the same idea scores the same for everyone.
- 75+ ideas earn a real validation sprint; 0–39 are usually features or hobbies.
- The score is a triage tool, not a success predictor — read the signal breakdown, not just the total.
- Personal-fit filters should be layered on top of the score, not baked into it.
Related reading
For acquisition cost expectations behind the score, read Finding your first 100 customers. For monetisation signal context, see SaaS pricing models explained. For market sizing depth, read TAM, SAM, and SOM explained.