GeoScript Methodology / Opportunity Score
Opportunity Score
A deterministic, versioned, explainable score summarising how attractive a business is as a prospect for an AI-visibility agency. Formula version: v1.
What the score answers
The Opportunity Score answers: How attractive is this business as a prospect for a GEO agency right now?
It is not a measure of how bad a business's AI visibility is. A business with very low visibility may still be a weak prospect if it operates in a low-value market, has no fixable issues, or has no marketing budget. Opportunity is about agency sellability, not raw visibility deficit.
Range, versioning, and determinism
Score range
0 – 100 (integer)
Higher = stronger prospect
Formula version
v1
Stored with every computed score
Determinism
Fully deterministic
Same inputs → same output, always
The formula version (v1) is stored alongside every opportunity score in the database. If GeoScript updates any component or weight, the version identifier increments. Historical scores remain valid under their original version and are not silently recomputed.
There is no LLM, no randomness, and no time-dependent value inside the scoring math itself. Inputs are derived from real observations; the formula is a pure function of those inputs.
The formula
Each of the eight components is scored 0–100, then multiplied by its weight. The weighted contributions sum to the raw total. A missing-evidence penalty is applied when observation count is low (see below). The result is rounded to the nearest integer, clamped to 0–100.
| Component | Weight |
|---|---|
| AI Visibility Gap | 25% |
| Commercial Value | 15% |
| Competitive Pressure | 15% |
| Fixability | 15% |
| Contactability | 10% |
| Marketing Maturity | 10% |
| Freshness | 5% |
| Evidence Confidence | 5% |
| Total | 100% |
Component definitions
AI Visibility Gap
25%How much AI visibility the business is missing relative to the market. Combines recommendation rate gap versus the market median, weighted share-of-voice deficit, coverage gap on high-commercial-intent prompts, and engine breadth deficit (present in only 1 of N engines is weighted more heavily than absent from 1 of N).
Commercial Value
15%Proxies for how attractive the business is as a customer. Combines industry CPC signal (normalised 0–1), business size proxy, review volume proxy (log-scaled), and website sophistication.
Competitive Pressure
15%How clear and compelling the competitor story is. Higher competitor dominance in AI results means a clearer pain narrative and, typically, a more motivated prospect. Combines competitor visibility advantage magnitude, count of direct competitors winning prompts, citation advantage held by competitors, and competitor engine breadth.
Fixability
15%Site-analysis signals indicating concrete, actionable improvement opportunities. Counts structured data gaps, weak entity clarity, poor answer coverage, and machine-readability issues. More concrete fixable issues raise this component.
Contactability
10%Ability to reach the business with contact signals currently available. Combines validated email (weighted highest), named decision maker, phone, LinkedIn presence, contact form, and enrichment confidence. A low Contactability sub-score does not disqualify an otherwise strong opportunity.
Marketing Maturity
10%Probability signal that the business is already buying marketing services. Signals include Google Ads pixel, Meta/Facebook pixel, retargeting infrastructure, CRM technology, analytics stack sophistication, SEO tooling, and marketing hiring signals.
Freshness
5%How recent the underlying evidence is. A freshness fraction of 1.0 means all evidence falls within its registered freshness window; 0.0 means all evidence has expired. Stale evidence lowers this component rather than being hidden.
Evidence Confidence
5%Aggregate confidence across the observation rows that produced the inputs, weighted by each component's contribution. Low confidence on a high-weight component (such as Visibility Gap) pulls this down more than low confidence on a lower-weight component.
Per-component explainability
Every computed score persists a full component breakdown: the raw sub-score for each component (0–100), the weighted contribution of each component to the total, the formula version, the observation count used, the missing-evidence penalty applied (if any), and the top contributing factor.
This breakdown is stored with the score record and surfaced in the product UI. No score is ever a single opaque number — any score card can expose the components that produced it.
Honest under-evidence behavior
When the number of observations used to derive the scoring inputs falls below the minimum threshold (currently 5 observations), GeoScript applies a missing-evidence penalty rather than computing a score as if all inputs were reliable.
The penalty scales linearly. At zero observations, the score is capped at approximately 40 out of 100, regardless of what the formula would otherwise return. At the minimum observation count, no penalty applies.
This means a business with very few observations will never receive an inflated score. Missing evidence lowers the score honestly — it is not treated as neutral or optimistic by default.
What the Opportunity Score does not measure
It is not purchase probability
The Opportunity Score is not a prediction of whether a business will become a customer. It ranks prospects by observable signals that correlate with agency sellability. Whether a specific business converts depends on factors GeoScript cannot observe: internal budget decisions, existing vendor relationships, and timing. Never treat the score as a guarantee of results.
It does not measure business health or revenue
Commercial Value uses publicly observable proxies (industry CPC, business size signals, review volume). GeoScript does not have access to actual revenue, profit margins, or internal budget data. The score reflects observable signals, not verified financials.
It does not measure AI-visibility quality
A high Opportunity Score means the business is an attractive prospect. It does not mean the business has poor AI visibility — a high score could come from strong Fixability or strong Competitive Pressure even if Visibility Gap is moderate.
Current limitations
Feature status: In Development
The Opportunity Score feature is currently in development and not yet generally available. The scoring function is implemented and deterministically tested, but it is not yet exposed to all users.
Several input signals are not yet fed by live data sources
Some scoring inputs currently use placeholder or default values while their upstream data sources are being built:
- Industry CPC signal — the normalised cost-per-click signal for the Commercial Value component is not yet drawn from a live data provider.
- Contact enrichment signals — contactability inputs such as validated email, named decision maker, and enrichment confidence are not yet fed by a live enrichment pipeline.
Scores computed while these signals are unavailable will reflect those gaps — they will not be inflated by optimistic defaults. The missing-evidence penalty and low-confidence signals will pull affected component scores down, not up.
Market median is a working estimate
The Visibility Gap component compares a business's recommendation rate against a market median. That median is currently set to a working estimate (0.5) rather than a calibrated value derived from real GeoScript data. It will be updated as observed data accumulates.