How an anonymous startup will be assessed

Site comparison versus full assessment: The website currently automates source text extraction, characteristic detection, portfolio analogues and a preliminary investor research shortlist. It does not perform every judgment below, verify commercial metrics, refresh investor terms, or generate calibrated funding probabilities. The full analyst workflow below describes a deeper follow-up assessment.

The assessment will estimate relative fit and investment plausibility, with confidence and reasons. It will not manufacture an acceptance percentage from a list of funded companies. We do not have the rejected applicants, admission-time features, complete cohort denominators or investor decisions required to calibrate such a probability. Even a very strong fit can have a low absolute chance at a selective program.

Input — no company name required

Describe the customer and painful problem; what the product does; the technical mechanism and differentiation; business model and pricing; stage and launch date; traction (users, revenue, growth, retention, pilots); founder experience and ability to build; geography and relocation constraints; previous funding and amount sought. For science/hardware, add experimental results, IP rights, regulatory pathway, manufacturing and technical milestones. Rough ranges are fine. Unknown fields remain unknown rather than counting against the startup.

I will not try to identify the company. I will use the description to compare business characteristics. If a distinctive description resembles a known portfolio company, I will flag possible overlap without asserting an identity or using its known fundraising result as a prediction.

Assessment sequence

  1. Eligibility first. Check geography, stage, current intake, incorporation, founder commitment, physical participation and sector restrictions. Distinguish a strict gate from a preference or typical pattern. A failed hard gate overrides portfolio similarity.
  2. Choose the correct route. Equity accelerator, venture studio, fellowship and credits/membership have different outcomes. Rank investment candidates separately from support opportunities. Treat SOSV and its subprogram as one investor relationship. Select a specific Techstars or Google program when possible.
  3. Retrieve analogues. Search the local company descriptions and sectors, then inspect the strongest two to five examples per candidate. Check the actual source before using a decisive analogue, especially YC mirror rows, historical 500 tags, SPC firm-level entries, or scarce HF0/PearX/Arc records. Compare customer, workflow, technology, go-to-market, stage and capital intensity; do not rely on an “AI” keyword alone.
  4. Judge the business. Evaluate founder-market fit, urgency, product evidence, retention/growth, distribution, differentiation, economics, scientific risk and venture-scale outcome. Weight these according to the program and stage. A scientific fellowship requires different evidence from a revenue-stage software accelerator.
  5. Look for disconfirming evidence. Existing close competitors may create conflicts or saturation. Similarity is not an endorsement. Legacy investments may no longer match today's thesis; later company success and current descriptions must not be mistaken for evidence available at admission.
  6. Report a ranked shortlist and the gaps. Use Strong fit / Plausible fit / Stretch / Currently ineligible, paired with High / Medium / Low evidence confidence. Missing private information often makes confidence low. These are analyst judgments, not measured acceptance-rate bins. Review all 19 entries, grouping parent/subprogram and support-only routes appropriately.

Output for each serious candidate

Field What you will receive
Fit and confidence A qualitative judgment with its assumptions
Why they might invest Specific connection to current thesis and startup evidence
Portfolio precedents Named companies, sources, similarities and meaningful differences
Why they might pass The strongest substantive objections and potential portfolio conflicts
What would change the judgment A concrete milestone or missing fact
Application angle A concise, truthful pitch tailored to that program
Timing and terms Rechecked official intake, eligibility and funding structure

The overall conclusion will separate an investable business, a good program fit, and an application likely to stand out. No single similarity score establishes all three. If you provide only an idea, the result will be a preliminary thesis-fit assessment; traction and team judgments will remain open.

The optional local scripts/match.py tool retrieves textual analogues. It runs offline against the saved dataset and does not send the brief elsewhere. Its numerical output measures word relevance, not funding likelihood, and must not be used to rank investors automatically.