How GritScore works
We analyse CV text against seven dimensions that distinguish people who thrive in startups from those who don't. Here's exactly what we measure and why.
The analysis process
CV text extraction
You paste plain text or upload a PDF. We extract the full text server-side using a document parser — nothing is stored after the session ends.
AI signal extraction
Claude reads the CV and identifies concrete evidence for each of the seven dimensions — citing specific roles, company names, and achievements. Vague language scores lower than concrete outcomes.
Dimension scoring
Each dimension is scored against a rubric tied to its maximum weight. Scores are anchored to CV evidence — the model cannot award points without citing specific supporting text.
Composite score & verdict
Scores are summed into a 0–100 composite. The verdict — STRONG FIT, UNCERTAIN, or RISK — is determined by score band and signal pattern, not just the number.
Interview probe generation
For each weak or unvalidated signal, the model generates a targeted interview question designed to probe that specific gap. You get 4–6 questions per CV, never generic.
The 7 scoring dimensions
Total score is out of 100. Each dimension has a maximum weight reflecting its relative importance for startup performance.
Ownership & Initiative
/20 ptsDid they drive outcomes independently, or wait to be directed? We look for self-started projects, unsolicited scope expansion, and evidence of decision-making without manager approval.
Ambiguity Tolerance
/18 ptsCan they operate when the problem isn't fully defined? We scan for roles with ill-defined briefs, pivots they navigated, and achievements made without a clear framework.
Measurable Impact
/17 ptsDid they move numbers, or just complete tasks? We weight quantified outcomes — revenue, growth rate, cost savings, time saved — over activity-based language.
Breadth of Responsibility
/15 ptsHave they worn multiple hats? Startup people do things outside their job title. We look for cross-functional work, covering gaps, and acting above their seniority.
Speed of Progression
/12 ptsDid they grow fast, or coast on tenure? Rapid promotion, expanding scope within a role, and early responsibility all signal the kind of trajectory startups attract.
Resourcefulness
/10 ptsWhat did they build with limited means? We look for bootstrapped outcomes — launching without a full team, using free tools creatively, delivering despite resource constraints.
Small Environment Experience
/8 ptsHave they worked where every person counted? Time at sub-50-person companies, early-stage roles, or scrappy teams scores positively. Big-company-only backgrounds score lower.
Score bands & verdicts
Strong evidence of startup-relevant traits across most dimensions. Worth prioritising in the interview process.
Mixed signals. Might thrive — but there are specific gaps worth probing before committing interview time.
CV evidence suggests the candidate is optimised for structured, large-company environments. Likely to struggle in a startup context.
Limitations & honest caveats
- ⚠CVs are curated. People write CVs to impress. The model can only score the evidence present — a well-written CV will score higher than an equally capable candidate who writes poorly.
- ⚠No personality data. GritScore reads career evidence, not personality. Two people with identical CVs may perform very differently in practice.
- ⚠Context matters. A 70/100 for a seed-stage role is different from a 70/100 for a Series B role. Use the score as a relative filter, not an absolute threshold.
- ⚠Use it to focus, not to decide. GritScore is an interview preparation tool. It surfaces where to probe — the interview is where you validate.
Free preview · Full report $19