How it works
Scoring & Intelligence
Opportunity fit scoring measures how well your existing application answers align with what a specific program asks for. AQUA computes significance scores, fit scores, composite scores, and program value signals — all preparation signals, not admissions predictions.
We measure signal, not truth. Every number on this platform is derived from a defined formula applied consistently to structured data. Scores are mathematically formulated internally — not editorial opinions, not committee judgments, not external rankings. The math is the same for everyone. What changes is the input.
The metrics
Five signals, each with a defined scope
Significance Score
How important a question is across the universe of programs. A high significance score means the question appears frequently, commands longer answers, and aligns with high-prestige themes.
Formula
asked_by_count x word_limit_weight x theme_prestige x universal_bonus
What it does not mean
This score says nothing about the quality of your answer. It measures the question's structural importance — how much the ecosystem cares about this category of question.
Fit Score
How well your current profile aligns to a specific program's DNA. It measures four dimensions: how much of the program's question surface you have covered, how closely your answer themes match what the program emphasizes, how well your stated criteria match the program's focus, and a quality signal derived from answer completeness.
Formula
(coverage_pct x 0.40) + (theme_alignment x 0.35) + (criteria_match x 0.15) + (quality_signal x 0.10)
What it does not mean
A fit score is not a prediction of acceptance. Programs admit founders for reasons that live outside any formula — timing, team dynamics, the partner reading your application that week. This score tells you where you are prepared relative to what the program measures.
Composite Score
A single opportunity signal that combines your fit with a program against that program's estimated value. Used to rank programs in the Hub when you are logged in. Programs where you fit well and that carry high opportunity value surface first.
Formula
fit_score x program_value_score / 100
What it does not mean
A composite score does not rank you against other founders. It ranks programs against each other, for you specifically. Two people viewing the same Hub see a different sort order.
Heat Score
ProvisionalA program desirability signal based on prestige markers, cohort size, follow-on funding rate, and structural indicators from the archive. Heat is used to surface programs when no personalized fit data is available.
Formula
f(prestige_weight, cohort_size, follow_on_rate)
What it does not mean
Heat is not real-time competition data. It reflects structural signals in our archive, not live applicant volume or this-cycle acceptance rates. Heat scores are marked provisional until we have sufficient longitudinal data to validate the model.
Program Value Score
An estimated opportunity value for a program, derived from brand weight, network quality, check size, and equity terms. Used as a multiplier in composite scoring. Higher value scores amplify good fit.
Formula
f(brand_weight, network_quality, check_size, equity_factor)
What it does not mean
This is not an objective ranking of programs against each other. It is a structural estimate, not an editorial opinion. A program with a lower value score is not a worse program — it may simply serve a different stage or sector where our current signals are thinner.
Signal depth
The same answer, two different scores
Two founders can give identical answers and receive different scores — because the meaning of words comes from their source.
A fit score is not purely about what you wrote. It is about what you wrote in the context of who wrote it — your profile, your history, the answers you have built over time, the themes your work touches.
This is the foundation of deeper signal analysis coming in future updates. As you build your answer bank, the system develops a more complete picture of your signal. A sparse profile produces shallow scores. A complete one produces scores that reflect real alignment.
The source-awareness layer is currently derived from profile completeness and theme coverage. More granular founder context signals are in development.
Boundaries
What we do not do
We do not decide who gets in
AQUA has no relationship with any program's admissions process. No score we surface influences any decision made by YC, Techstars, NSF, or any other program.
We do not rank founders against each other
Scores are personal. Your composite score tells you how you fit a program — not how you compare to other people applying to that program. We never expose inter-founder comparisons.
We do not sell answers
Your answer bank is row-level-secured to your account. We do not aggregate, train on, or surface individual answers to anyone. Your capital stays yours.
We are not the arbiters of truth
The decision belongs to the programs, not us. We surface signal that helps you prepare. What programs do with what you submit is entirely outside our scope — and we intend to keep it that way.
In development
Signals we are building toward
MoatScore
A commitment conservation signal. Measures how consistently a founder's stated direction holds across applications over time. Applicants who drift lose signal — those who maintain a clear thesis gain it.
FundScore
An answer fidelity extension to the fit model. Evaluates whether the quality and specificity of answers matches the signal weight the question carries — not just that the question was answered.
Both extend the current signal framework. No timelines — they ship when the data supports them.