How the SFRI works
SFRI keeps four answers deliberately apart - internal readiness, market evidence, confidence, and private founder signals - so you can see what is strong and what needs attention, without one number hiding another.
The SFRI does not produce a single magic number. The Internal Readiness Score (IRS) measures how structurally ready the company itself is: money, people, learning speed, scaffolding. The Market Evidence Gate (MES) checks whether the market is actually pulling, on a strict Weak / Emerging / Strong ladder that internal strength cannot compensate for. The Confidence Score (CS) reports how well your answers cover the model - evidence adequacy, not a probability of success. And a private founder channel holds your personal well-being signals, visible only to you. Every score comes from transparent, rule-based formulas: the same answers always produce the same result, and no AI decides anything - an AI only turns your computed results into readable sentences.
Have three things at hand: your current cash balance, your last three months of burn and revenue, and a rough sense of your top-3 customer revenue shares. You can enter amounts in EUR or HUF - HUF is converted once, at the official MNB mid-rate of the day, and the rate is stored with your session. Everything else you can answer from memory. Under an hour, no login, results on screen immediately.
The four internal dimensions
The Internal Readiness Score (IRS) is the stage-weighted blend of four dimensions, each answering one plain question:
“Can you survive a shock?”
Built from your cash runway, a stress-tested runway (costs up 20%, revenue down 30% at once), revenue concentration in your top customers, how predictably you hit your own spending plan, and whether the critical basics - financial tracking, customer list, code repository, agreements - actually exist.
“Can the human system hold?”
Built from how safely your team speaks up, how much friction exists between co-founders, your own well-being and work energy over the last two weeks, and whether key people have stayed over the last 12 months. Startups run on this system long before they run on process.
“Can you learn and act quickly?”
Built from how many deliberate experiments you ran in 90 days, how many ended in a clear decision, how fast you responded to the last serious negative signal, and how long your last significant strategic decision took.
“Can you scale with support and structure?”
Built from how deliberately you build and use partner relationships, how embedded you are in the ecosystem around you - mentors, investors, peers, institutions - and whether governance and responsible-scaling foundations are in place for growth to land on.
The Market Evidence Gate
A startup can be internally strong and still market-unproven. SFRI keeps those signals separate. The gate asks stage-appropriate questions - problem-discovery depth before revenue; retention, growth and customer pull at Seed; revenue retention and sales efficiency later - and grades them on a strict ladder. The gate is non-compensatory: Strong requires every Strong criterion at once. And honest "I don't measure this" answers count as answers, scored conservatively - never punished as missing data.
What comes out
0-100, banded High (65+), Medium (40-64) or Low (<40). Below its evidence threshold a dimension says Insufficient Evidence - no score is ever invented.
Weak, Emerging or Strong. Reported next to the IRS, never blended into it.
How adequately your answers cover the model, labelled High / Medium / Low. Evidence adequacy - not a probability.
The combined readout of IRS band and MES tier. Nine possibilities, from Investment-Ready to Critical.
Your exact score and band, e.g. "3.7 / 5 - Recovers steadily from setbacks". Outside every other score, by design.
Continuity risks, like runway under 3 months, that cannot be averaged away and override the headline profile.
A strengths-first summary composed from your rule-based results, plus a concrete list of what to measure next.
The full archetype map
Your archetype is where your IRS band meets your MES tier - nine cells, plus two outcomes that sit outside the matrix.
both the organisation and the market confirm readiness.
internal strength with early market signals; positioned for acceleration.
a strong organisation building something the market has not validated.
the market wants it, but the organisation may break under demand.
market interest exists, but organisational capacity may not sustain delivery.
neither internal systems nor market demand are at threshold.
strong market pull into a structurally unsound organisation.
demand exists, but the organisation lacks the basic infrastructure to respond.
no internal capacity and no market validation; fundamental rethinking required.
a venture-continuity red flag overrides every cell until it is addressed.
below the evidence threshold, no archetype is asserted.
Technical detail (optional)
Stage weights and bands
Your stage — Pre-Seed, Seed or Series A+ — is classified from capital raised, revenue, team size and product status (majority vote; ties resolve to the lower stage). The Internal Readiness Score (IRS) weights differ by stage:
| Stage | OR | FTH | AE | ELRS |
|---|---|---|---|---|
| Pre-Seed | 35% | 35% | 20% | 10% |
| Seed | 30% | 25% | 30% | 15% |
| Series A+ | 25% | 20% | 25% | 30% |
How the Confidence Score (CS) is built
Each answer carries an evidence class: later-verifiable factual inputs count fully, psychological self-reports count at 70%. Per component, confidence = completeness × average evidence quality; the market evidence component also loses 10% per open consistency alert. The overall Confidence Score (CS) is the mean of the five components, labelled High (0.85+), Medium (0.70–0.84) or Low. It describes evidence adequacy — never a probability of success.
Exemptions vs. missing data
Structural exemptions — a solo founder skipping co-founder questions, a very young company skipping the spending-plan check — shrink what the model expects and cost nothing. Skipped answers are different: a dimension is only scored when enough of its planned metrics are present, including its core metric. Otherwise it reads “insufficient evidence”, no score is invented, and the Evidence Gap Report tells you exactly what to bring next time. A solo founder's team questions and a pre-revenue startup's customer-concentration question are structural exemptions too - the model expects less, it does not punish.
Well-being (WHO-5) and work exhaustion (CBI) readings appear only on your own results screen - never in the PDF - and none of them can cap or flag your venture's result. Your resilience result is the one personal signal that appears in the profile, and you are told before answering those six questions.