Idea Validation, in depth
12 lenses.
One honest verdict.
PRISM is Idea Validation's framework — the layer that sits on top of QUAD specifically to answer "should I build this?" This page goes one level deeper than the shared methodology hub: the 12 lenses, how evidence is graded, and exactly how a KILL is reached.
The 12 lenses
Every idea runs through all 12, every time
The same 12 lenses, in the same order, for every idea submitted — so one verdict can be meaningfully compared against another, and no idea gets an easier or harder read depending on how it's framed.
Evidence grading
Every quantitative claim earns a grade, A–E
Within each lens, any specific number a founder submits — a market size, a claimed cost saving, a churn benchmark — is checked against available evidence and graded conceptually as follows:
The grade travels with the claim into your report. This is the same rubric Pre-Launch Verification's Audit uses on pitch-deck claims — one evidence standard across products, not a different bar for each.
The gate logic
How a KILL is actually reached
PRISM doesn't average 12 lens scores into one composite number and call whatever crosses a threshold a "GO." A small number of lenses — most commonly Competition & Moat and Business Model & Unit Economics — carry a hard kill criterion attached. If a Gate lens fails, the overall verdict is KILL (or PIVOT, if the failure is fixable with a clear change) regardless of how strong the other 11 lenses score.
This mirrors how a real, skeptical investor actually reads a deck: a beautiful go-to-market plan doesn't rescue an idea with no defensible moat, and a large addressable market doesn't rescue unit economics that don't work. Gate failures are foundational, not cosmetic — which is exactly why they override the composite score rather than being averaged into it.
Stage-awareness matters too: a pre-seed idea and a Series A-ready idea aren't held to the same evidentiary bar on every lens — Team & Capability, for instance, is scored differently for a solo pre-seed founder than for a Series A team expected to already have a full bench. What doesn't change by stage: the Gate lenses. A moat that doesn't exist, or economics that don't work, are a KILL at any stage.
See a full worked example — all 12 lenses scored, the exact Gate failures, and the signed verdict — on the sample Idea Validation report.
What we don't publish, and why
Enough to trust it, not enough to game it
We don't publish the exact weighting between lenses, the numeric scoring thresholds that separate a CONDITIONAL GO from a PIVOT, or the method used to generate the 60+ stress-test scenarios each idea is run against. Publishing those would let anyone reverse-engineer a submission specifically crafted to score well rather than one that's actually sound — the opposite of what a validation framework is for. What we do publish — the 12 lenses, the A–E grading standard, and which lenses carry hard gates — is everything you need to understand why your idea got the verdict it did.
Run it on your idea
Which verdict would your idea earn?
Start with a free Pulse — a 5-question, 15-minute diagnostic. No card required.