Methodology · version of October 2, 2026

How the assessment is calculated, and its limits.

Principles

  • Only your figures. No value is inferred or guessed: missing data shows “not provided”, never a silent estimate.
  • Published methods. Each verdict cites its primary source, checked at the source, with its published critiques.
  • No overall score. No study justifies weighting cash, product-market fit and organization into a single score. Academic research on why startups fail remains thin — for example, a 2025 peer-reviewed study covers 40 startups (Journal of Innovation and Entrepreneurship).
  • Privacy. The calculation runs in your browser: your figures are neither sent nor saved.

Order of priority

When several risks appear, they are ranked as follows: cash, product-market fit, capital efficiency, organization. This is reasoning, not a statistic: running out of cash is the most cited final cause of startup shutdowns, and insufficient product-market fit the most cited root cause (CB Insights, March 2026, 431 funded startups shut down since 2023). These figures rely on public post-mortems and declared causes: we only use them for the order, not as a measure.

The suggested C-Level role matches the usual scope of the position for the priority risk (cash and efficiency: CFO; product-market fit: CEO; organization: CHRO), among the 10 roles offered at launch. No rigorous study measures when a startup needs a fractional executive: this is reasoning, shown as such.

Cash: default alive / default dead

What it measures

If your spending stays constant and your revenue keeps growing at the same pace as in recent months, do you reach profitability before running out of cash?

Calculation

Month-by-month projection, over 10 years at most: revenue of the month = current revenue × (1 + monthly growth)^month; cash decreases by the month’s burn. “Default alive” if revenue exceeds spending before cash goes below zero, “default dead” otherwise. Gross margin (optional): the spending entered is split into fixed costs + direct costs (revenue × (1 − margin)); today’s burn is unchanged, but each extra euro of revenue only covers its margin. This is our own extension — Paul Graham reasons in revenue, which amounts to assuming a 100% margin, an optimistic assumption outside software, applied by default and flagged on screen. Without a growth rate, only the cash runway at constant revenue is given. Without revenue, growth doesn’t apply: the verdict is mechanically “default dead”, and the result explains it by highlighting the cash runway available.

How we use it

It’s a calculation on your own figures, not an external statistic: it doesn’t depend on any sample. Its limit is its assumption — constant spending, unchanged growth — which we show with every result. It’s a simplified projection, not a forecast.

Capital efficiency: burn multiple

What it measures

How much cash you burn for every euro of new annual recurring revenue.

Calculation

Burn multiple = annual net burn ÷ new net ARR over 12 months. Annual burn is the actual net burn of the last 12 months if provided; otherwise it’s estimated from the current monthly burn × 12, an approximation flagged on screen (the two periods then differ). Published benchmarks: < 1 amazing, 1 to 1.5 great, 1.5 to 2 good, 2 to 3 suspect, > 3 bad. Calculated only for recurring revenue models.

Published critiques

  • The scale assumes a next fundraise will happen, plus a 75–80% gross margin and net retention ≥ 100%. A good burn multiple is still burn: the cash-out date matters more if the raise isn’t certain. (Jason Lemkin, SaaStr)

How we use it

The formula is reliable; the scale is an investor’s heuristic, not the result of a study. It’s always shown next to the cash verdict, never on its own.

Product-market fit: Sean Ellis test

What it measures

The share of users who would be “very disappointed” if they could no longer use your product.

Calculation

Question: “How would you feel if you could no longer use the product?” (very disappointed / somewhat disappointed / not disappointed). Benchmark: at least 40% “very disappointed”. Below 40 responses, we don’t conclude.

Published critiques

How we use it

We keep the question, simple and useful, but present the 40% as a benchmark not scientifically validated, to complete with retention.

Team growth before market validation

What it measures

Whether your headcount grew over 12 months while your market isn’t validated yet (“Discovery” or “Validation” stage, or Sean Ellis test below the benchmark).

Calculation

Signal = current headcount > headcount 12 months ago AND market not validated. No numeric threshold is applied: it’s an observation on your figures, not a score.

Published critiques

  • Survivorship bias (only companies still alive were surveyed), applying global trends to individual cases, multiple statistical testing. The Startup Genome team replied in the same article. (Jason Cohen, juin 2011)

How we use it

We only keep the principle (don’t grow before validating your market) and the stage definitions, based on milestones rather than funding. The report’s percentages (2011 data, self-reported, correlational) aren’t shown.

Tech: software delivery performance

What it measures

We calculate nothing on this axis.

Calculation

We point to the official DORA Quick Check (free, no sign-up), which compares your team with industry benchmarks on the DORA metrics.

Published critiques

How we use it

It’s the most solid source on this list (multi-year, published research). Rather than copying its thresholds, we point to its authors’ tool.

The quantified levers (“what would change the verdict”)

The assessment doesn’t stop at a verdict: it calculates what would change it. No outside data is involved — we replay exactly the same cash projection, varying one parameter at a time, and use a binary search to find the smallest value that makes the startup “default alive”.

  • Default dead : minimum monthly revenue growth (tested up to 100% / month), spending cut needed, or extra cash needed. If revenue doesn’t grow, no cash injection is enough: the assessment says so instead of showing an amount.
  • Default alive : the floor growth below which the startup would tip into default dead.
  • Without a growth rate : the growth that would be needed, to compare with actual growth.
  • Without revenue : the approximate month cash runs out and the monthly revenue that would cover current spending (spending ÷ gross margin).
  • Burn multiple of 2 or more : the new net ARR (at unchanged burn) or the maximum annual burn (at unchanged ARR) that brings the ratio below 2, the threshold below which it leaves the “suspect” and “bad” bands of David Sacks’ scale.

These are arithmetic thresholds, not recommended targets: they hold “all else being equal”, and combining several levers takes less effort on each. Thresholds are rounded on the cautious side (to the nearest €100 for spending, €1,000 for cash, ARR and burn, 0.1 point for growth), so they remain sufficient.

What the assessment doesn’t measure (yet)

  • User retention, though recommended alongside the Sean Ellis test.
  • Unit economics beyond the burn multiple (acquisition cost, customer lifetime value, margin).
  • The French or European context: all the sources used are American and venture-capital oriented.

Found an error, have a more solid source to suggest? contact@byc-level.com