Profile illustrative
Dimensions weight · score
click a dimension for its 3 criteria
VC method inputs
context
Hardware startups fail differently from software startups. In the VC method, technical risk only enters through the investor's gut-feel return multiple, so a hardware startup's valuation depends on how risky it feels, not on how mature it is.
the model
- 7 dimensions with 3 criteria each, every criterion scored 1 to 5 against written score anchors
- Weighted total: technology readiness and verification 20% each, integration, reliability and manufacturing 15% each, regulatory 10%, supply chain 5%
- Red-flag rule: any single criterion at 1 is reported separately, so one fatal gap cannot be averaged away
- Risk classes: I from 4.0, II from 3.0, III from 2.0, IV below 2.0
link to the VC method
Each risk class sets a surcharge on the required return multiple: none for class I, 20 to 30% for class II, 50 to 75% for class III, and no investment for class IV. In the worked example, a score of 2.40 lowers the pre-money valuation from €9.86M to €5.29M, a 46.3% drop that would otherwise stay hidden in the investor's intuition.
validation
- Expert interviews with investors from three venture funds, including a corporate VC
- A case study with a pre-seed hardware startup
- A consistency check with 10 AI agents in different investor roles over 3 runs: 28 valid assessments of the same fictional startup landed within a narrow band around 2.4
result
Thesis and colloquium both graded 1.0. The surcharges are orientation corridors, not market-calibrated values, which the thesis states openly.
skills
technical due diligence · venture capital · vc method · mcda · research design · llm evaluation