01Principles
Evidence before verdicts
SaaSift does not ask a model to invent a single market score. It collects bounded signals, preserves their provenance, normalizes comparable inputs to a 0–100 scale, and applies published weights. The same validated inputs and scoring version produce the same result.
Deterministic
Fixed inputs produce a fixed score.
Versioned
Formula changes receive a new scoring version.
Inspectable
Coverage and evidence remain visible beside the score.
02Evidence types
Facts and interpretations stay separate
Observed
Direct market evidence
A provider returned keyword volume, or a pricing page exposed a public monthly price.
Derived
Transparent calculation
Organic traffic divided by authority, or the share of low-difficulty keywords in a validated set.
Inferred
Bounded interpretation
An AI-assisted assessment of product scope, paired with an explicit confidence value and source context.
03Opportunity score
Six weighted dimensions
Each dimension is calculated from smaller components. Available dimension scores are combined using the weights below; score coverage is calculated separately.
01
Demand
Is there measurable demand for this problem?
Search volume, existing traffic, keyword breadth, non-brand coverage, commercial intent, and tool intent.
02
Commercial
Does the market show evidence that buyers pay?
Public pricing, payment evidence, comparable CPC, paid-feature clarity, and recurring or credit-based models.
03
SEO
Can a focused product acquire relevant organic traffic?
Organic share, traffic-to-authority efficiency, low-difficulty keywords, weak top pages, and programmatic potential.
04
Competition
Are the current search results vulnerable?
Low-authority results, intent mismatch, weak product coverage, link weakness, brand diversity, and stale pages.
05
Buildability
Can a small team ship and maintain a useful first version?
Frontend, backend, infrastructure, APIs, data, AI/GPU, compliance, and ongoing maintenance complexity.
06
Trend
Is market momentum improving or deteriorating?
Versioned traffic and market snapshots when sufficient comparable history is available.
04Missing data
Unknown is not zero
A provider outage, unsupported country, or unavailable metric is represented as null. It is excluded from the weighted numerator and denominator, while coverage falls to show how much of the intended model was actually available.
Demand: 80 × 20% weight
All other dimensions: unavailable
Score: 80 · Coverage: 20% · Verdict: Insufficient Data
This prevents unavailable data from looking like poor market performance while still blocking high-confidence conclusions from thin evidence.
05Verdicts
Scores are filtered through decision rules
Hard filters take precedence
- Coverage below 65% → Insufficient Data
- Buildability below 30 → Poor Build Fit
- Commercial below 25 → Low Monetization
- Competition below 20 → Too Competitive
- Confidence below 50 → Needs More Evidence
06Limitations
A research aid, not a promise
Search and traffic providers estimate parts of the market and can have gaps. Public websites do not reveal private revenue, retention, margins, customer satisfaction, or founder advantages. SaaSift does not present modeled revenue as reported company revenue.
A high score means the available evidence supports deeper validation. It does not guarantee demand, ranking performance, product-market fit, or commercial success.