Data Quality
Understanding Darwin's data quality scoring system
Data Quality
Darwin provides a data quality score to help you understand the confidence level of your biodiversity assessment.
Quality Score Overview
The score reflects the mix of data types used in your assessment:
| Score | Label | Primary Data Type | Confidence |
|---|---|---|---|
| A | Excellent | Pressures | Highest |
| B | Good | Commodity | High |
| C | Moderate | Product | Medium |
| D | Indicative | Financial | Lower |
How Scores Are Calculated
The quality score is computed based on the weighted contribution of each data type to your total footprint:
Quality Score = Σ (Data Type Weight × Data Type Quality Factor)
Quality Factors
| Data Type | Quality Factor | Rationale |
|---|---|---|
| Pressures | 1.0 | Direct measurement, no modeling assumptions |
| Commodity | 0.75 | Specific raw materials, limited modeling |
| Product | 0.5 | Product-level averages, moderate uncertainty |
| Financial | 0.25 | Sector averages, highest uncertainty |
Score Interpretation
Score A (Excellent)
Your assessment is primarily based on direct pressure data. Results have the highest confidence and are suitable for detailed reporting.
Typical profile: Organizations with direct emissions measurements, detailed water and land use data.
Score B (Good)
Your assessment uses significant commodity-level data. Results are reliable for most reporting purposes.
Typical profile: Organizations tracking raw material procurement in physical units.
Score C (Moderate)
Your assessment relies on product-level data. Results provide a good estimate but could be refined.
Typical profile: Organizations with product procurement data but limited commodity detail.
Score D (Indicative)
Your assessment is primarily based on financial data. Results are indicative estimates useful for prioritization but may have significant uncertainty.
Typical profile: Organizations starting their biodiversity journey with spend-based analysis.
Improving Your Score
To improve data quality:
- Identify hotspots: Focus on entities/products with highest footprint
- Collect better data: Move from financial → product → commodity → pressure
- Prioritize impact: Better data on high-impact areas improves overall quality
You don't need perfect data everywhere. Focus on the 20% of activities driving 80% of your footprint.
Dashboard View
In the Compare view, you can see:
- Data mix breakdown: Pie chart showing % by data type
- Average quality score: Weighted score across all data points
- Quality by entity: Score breakdown by organization unit
Best Practices
Starting Out
- Begin with financial data for a quick overview
- Quality score D is normal and useful for initial screening
Improving Assessment
- Request physical data from key suppliers
- Collect direct measurements for operational sites
- Use commodity data for high-impact procurement categories
Reporting
- Always disclose your data quality score in reports
- Explain the mix of data sources used
- Document plans for data improvement
Last updated today