Methodology

How the Data is Created

Our pipeline from raw sources to structured trade intelligence.

01

Source Collection

We continuously ingest content from thousands of global sources — news wires, government trade bulletins, port authority notices, commodity exchanges, and regulatory filings across 40+ countries.

02

Event Detection

Natural language processing models scan incoming content to identify trade-relevant events — disruptions, policy changes, infrastructure developments, supply shifts, and market restorations.

03

Classification

Each detected event is automatically classified across multiple dimensions: event type, affected commodities, geographic region, impacted trade routes, and temporal status.

04

Severity Scoring

Events are assigned a severity score on a bidirectional scale from -4.0 (major disruption) to +4.0 (significant restoration), reflecting both the magnitude and direction of trade impact.

05

Enrichment

Structured records are enriched with trade route mapping, HS commodity codes, geographic coordinates, and cross-references to related events for full contextual depth.

06

Delivery

Final records are published daily via our API as structured JSON, ready for integration into trading systems, risk dashboards, research platforms, and supply chain tools.

Severity Scale

Our bidirectional severity score captures both disruptions and restorations on a single continuous scale.

-4.0 Critical disruption — major supply chain breakdown, port closure, or trade embargo
-3.0 Severe disruption — significant operational impact, potential cascading effects
-2.0 Moderate disruption — localised impact, partial service reduction
-1.0 Minor disruption — limited scope, manageable delays or restrictions
+1.0 Minor positive — incremental improvement, new trade agreement, or capacity expansion
+2.0 Moderate positive — resumed operations, lifted restrictions, or recovered routes
+3.0 Strong positive — major trade corridor reopening or significant policy breakthrough
+4.0 Critical restoration — full normalisation of major disrupted trade flow

Event Classification

Every event is tagged with a primary type reflecting the nature of the disruption or development.

Strike Supply Cutoff Policy Import Export Infrastructure Production

Commodity Coverage

We track disruptions across 18 major commodity categories spanning energy, agriculture, metals, manufacturing, and logistics.

Crude Oil Natural Gas Semiconductors Steel Iron Ore Rice Sugar Livestock & Meat Shipping & Logistics Electricity Agricultural Products Finance & Trade Policy

Data Quality

We maintain data integrity through multiple layers:

  • Duplicate detection and event deduplication across sources
  • Confidence scoring on NLP classifications
  • Temporal tracking — events are tagged as upcoming, happening, or resolved
  • Daily dataset refresh with full historical backfill
  • Human-in-the-loop review for edge cases and high-severity events

Output Format

Each record is delivered as a structured JSON object with 21 fields including severity score, event type, commodity tag, trade routes, geographic coordinates, HS chapter codes, and temporal metadata. See the data section on our homepage for a sample record.