Methodology

How Scores Work

A transparent, reproducible scoring system built on authoritative data. No black boxes. Every sub-score traces to a named source.

How Composite Scores Work

Every Axiom Locus composite score is a weighted average of 8 independent signal groups, each scored on a 0-100 scale. The signal groups cover different dimensions of location quality: commercial health, population trends, demographics, economic strength, development activity, infrastructure, safety, and amenity demand.

Each signal group is itself a weighted average of 3-6 sub-scores, each derived from a specific, named data source. This two-level aggregation ensures that no single data point dominates the final score.

Composite = sum(group_score[i] * weight[i]) / sum(weight[i])

Signal Group Weights

Default weights reflect the general importance of each signal group for CRE site selection. Weights can be customized through scoring profiles optimized for specific use cases like QSR, self-storage, retail, or office.

Signal GroupGeneralQSRStorageRetailOffice
Business Vitality
20%15%10%25%8%
Population Momentum
15%10%25%10%10%
Demographics
12%20%10%12%18%
Economic Strength
15%10%20%12%22%
Development Pipeline
12%8%15%8%10%
Accessibility & Infrastructure
10%12%5%18%18%
Safety & Environment
8%5%5%5%4%
Amenity & Demand
8%20%10%10%10%

Sub-Score Calculation

Each signal group aggregates 3-6 sub-scores using a weighted internal average. Sub-score weights are fixed (not affected by profile selection) and reflect the relative importance and reliability of each data source within its group.

For example, Business Vitality combines: Net Business Openings (25%), Category Diversity (20%), Rating Trajectory (20%), Business Density (20%), and Review Volume Growth (15%).

group_score = sum(sub_score[j] * sub_weight[j]) / sum(sub_weight[j])

Normalization

All sub-scores are normalized to a 0-100 scale using fixed-range linear or logarithmic scaling. Each metric has a defined min/max range based on national benchmarks. Linear scaling maps the raw value proportionally within that range; log scaling is used for metrics with wide distributions (e.g., population density, permit values) to prevent outliers from compressing the useful range.

Fixed-range normalization keeps scores comparable across metros and stable over time. A median income of $80K maps to the same score whether the location is in Nashville or San Francisco, making cross-metro comparisons meaningful.

Scores range from 0 to 100. Values below the range floor clamp to 0; values above the ceiling clamp to 100. Scores above 70 are considered strong; below 30 indicates significant weakness.

Confidence

Every score includes a confidence value from 0.0 to 1.0, calculated as:

confidence = sources_available / sources_possible

A confidence of 1.0 means all expected data sources returned data for this location. Lower confidence indicates some sources were unavailable, which can happen when:

  • City-specific sources (permits, crime data) don't cover the location
  • API rate limits or temporary outages prevent data retrieval
  • The location is in a rural area with sparse data coverage
  • Commercial data sources have gaps in their coverage

The composite confidence is a weighted average of group-level confidences, using the same weights as the score itself.

Scoring Profiles

Five built-in scoring profiles adjust signal group weights for specific CRE use cases. The underlying data and sub-score calculations remain the same; only the group-level weights change.

General

Balanced weights for general CRE evaluation. Good starting point for multi-use analysis.

Business: 20%Pop Momentum: 15%Demographics: 12%Economic: 15%Dev Pipeline: 12%Accessibility: 10%Safety & Env: 8%Amenity: 8%

QSR / Fast-Casual

Optimized for QSR and fast-casual restaurants. Emphasizes demographics, amenity demand, and accessibility over development pipeline.

Business: 15%Pop Momentum: 10%Demographics: 20%Economic: 10%Dev Pipeline: 8%Accessibility: 12%Safety & Env: 5%Amenity: 20%

Self-Storage

Tailored for self-storage facility siting. Heavily weights population momentum and economic strength — the primary demand drivers.

Business: 10%Pop Momentum: 25%Demographics: 10%Economic: 20%Dev Pipeline: 15%Accessibility: 5%Safety & Env: 5%Amenity: 10%

Retail

Designed for retail site selection. Prioritizes business vitality and accessibility, the key foot-traffic and co-tenancy signals.

Business: 25%Pop Momentum: 10%Demographics: 12%Economic: 12%Dev Pipeline: 8%Accessibility: 18%Safety & Env: 5%Amenity: 10%

Office

Built for office CRE decisions. Emphasizes economic strength, demographics, and accessibility for workforce-oriented locations.

Business: 8%Pop Momentum: 10%Demographics: 18%Economic: 22%Dev Pipeline: 10%Accessibility: 18%Safety & Env: 4%Amenity: 10%

Transparency

Every sub-score in Axiom Locus traces to a specific, named, authoritative data source. There are no proprietary black-box algorithms or opaque model outputs. The scoring methodology is fully documented here and in our signal group pages.

Our data sources include federal government datasets (Census, BLS, FEMA, EPA), municipal open data portals, and well-known commercial APIs (Google, TomTom, Adzuna). See our full source catalog for details.

We believe transparency builds trust. If a score doesn't match your on-the-ground experience, you can drill into the sub-scores and data sources to understand why.

Validation

We continuously validate our scoring model against historical outcomes. Our backtesting methodology compares predicted trends against actual observed changes in property values, lease rates, and business activity.

View our backtesting results and methodology at /backtest.