Processing Score Deep Dive
The complete technical guide to how we calculate Processing Scores, from ingredient parsing to penalty application, based on analysis of 1.97 million products.
The Science of Processing Measurement
The Processing Score (PS) quantifies the degree of industrial transformation a food has undergone from its original whole food state. Unlike simple ingredient counts, our system analyzes the type, purpose, and industrial nature of each component.
Core Principles
- • Additive approach: Scores only increase, never decrease
- • Uncapped scale: No maximum limit reflects reality of processing
- • Evidence-based penalties: Based on scientific literature
- • Transparent calculation: Every penalty is traceable
- • Ingredient-focused: What's in it matters more than marketing claims
Minimally processed
Ultra-processed
Products analyzed
Step 1: Base Score Assignment
Every product starts with a base score determined by its fundamental complexity:
Single Ingredient
Pure, unmodified whole foods with no additives or processing agents.
Examples: Apple, egg, milk, honey, olive oil, raw nuts
2-3 Simple Ingredients
Minimal combinations using kitchen-available ingredients.
Examples: Peanut butter (peanuts, salt), cheese (milk, salt, cultures)
Basic Processing
Traditional preservation methods, fermentation, or simple cooking.
Examples: Bread, yogurt, pickles, canned tomatoes, frozen vegetables
Standard Processed
Industrial production with multiple additives for function and preservation.
Examples: Crackers, cereal, pasta sauce, salad dressing
Special Cases
- • Water: PS = 0 (reference point)
- • Salt/Sugar: PS = 1.0 (single ingredient despite being refined)
- • Fermented products: Lower base score due to beneficial processing
- • Infant formula: High base score (5.0) due to extensive modification
Step 2: Processing Penalties
After the base score, we add penalties for specific industrial ingredients and techniques:
| Ingredient Type | Penalty | Rationale |
|---|---|---|
| Artificial Colors | +2.0 | Synthetic, linked to hyperactivity |
| Artificial Flavors | +2.0 | Lab-created, masks poor quality |
| Hydrogenated Oils | +1.5 | Industrial process, trans fats |
| High Fructose Corn Syrup | +1.5 | Enzymatically modified sugar |
| BHA/BHT/TBHQ | +1.5 | Petroleum-derived preservatives |
| Artificial Sweeteners | +1.2 | Synthetic sugar substitutes |
| Modified Starch | +0.8 | Chemically altered structure |
| Emulsifiers (industrial) | +0.6 | Synthetic texture modifiers |
| Natural Flavors | +0.3 | Extracted, concentrated compounds |
| Citric Acid | +0.1 | Natural preservative |
Cumulative Effect
Penalties are additive. A product with multiple problematic ingredients accumulates all applicable penalties:
Example: Flavored Chips
Base (5.0) + Artificial flavor (2.0) + MSG (0.6) + BHT (1.5) + Modified starch (0.8) = PS: 9.9
Step 3: Ingredient Classification System
Our system categorizes 270,000 unique ingredients into 84 functional buckets:
Major Categories
- • Additives (12 types): Colors, flavors, preservatives
- • Sweeteners (8 types): Natural, artificial, sugar alcohols
- • Fats/Oils (15 types): Natural, modified, synthetic
- • Proteins (10 types): Whole, isolated, hydrolyzed
- • Starches (7 types): Native, modified, resistant
Processing Indicators
- • "Modified": Chemical alteration (+0.8)
- • "Hydrogenated": Industrial saturation (+1.5)
- • "Isolated": Extracted component (+0.6)
- • "Hydrolyzed": Broken down (+0.5)
- • "Enriched": Synthetic addition (+0.3)
Normalization Process
We normalize ingredient variations to identify true components:
Advanced Calculation Examples
Example 1: "Healthy" Granola Bar
Ingredients: Oats, brown rice syrup, chocolate chips (sugar, cocoa, soy lecithin, vanilla), almonds, natural flavor, salt, vitamin E
Despite "healthy" marketing, this scores as processed due to industrial ingredients.
Example 2: Diet Soda
Ingredients: Carbonated water, caramel color, phosphoric acid, aspartame, potassium benzoate, natural flavors, citric acid, caffeine
Zero calories doesn't mean zero processing - heavily modified with chemicals.
Example 3: Plain Greek Yogurt
Ingredients: Cultured pasteurized milk, live active cultures (S. thermophilus, L. bulgaricus)
Beneficial fermentation actually reduces the score - minimal processing.
Handling Edge Cases
Organic Products
Organic certification doesn't affect Processing Score. Organic cookies with 30 ingredients score the same as conventional versions. We evaluate what's in it, not how it's grown.
Fortified Foods
Added vitamins and minerals receive small penalties (+0.2) as they indicate processing, but we recognize their nutritional benefit in the Nutrition Score.
Example: Fortified milk (PS: 1.7) vs Regular milk (PS: 1.0)
Restaurant Items
Fast food and restaurant items often lack complete ingredient lists. We use category averages and known preparation methods to estimate scores.
Example: McDonald's Big Mac (estimated PS: 14.5) based on 40+ known ingredients
Validation and Accuracy
Quality Checks
- ✓ Manual review of 10,000 products
- ✓ Cross-validation with NOVA system
- ✓ Expert nutritionist consultation
- ✓ Continuous algorithm refinement
- ✓ User feedback integration
Data Sources
- • USDA FoodData Central
- • OpenFoodFacts database
- • Manufacturer websites
- • Product packaging scans
- • Regulatory filings
Accuracy Metrics
94%
Ingredient parsing accuracy
98%
Category classification
±0.5
Average score variance
Limitations and Considerations
Processing ≠ Unhealthy
Some processing improves safety (pasteurization), nutrition (fortification), or accessibility (freezing). The PS measures degree of processing, not health impact.
Ingredient Order Matters
We don't account for ingredient proportions. A product with HFCS as the last ingredient gets the same penalty as one where it's the second ingredient.
Home Cooking Gap
We can't distinguish between industrial and home preparation. A homemade cake would score similarly to a packaged one with the same ingredients.
Regional Variations
Products may have different formulations in different regions. Our scores reflect the most common U.S. formulation.
Future Enhancements
🚀 In Development
- • Proportion-weighted penalties
- • Processing method detection
- • Supply chain analysis
- • Environmental impact factors
- • Packaging assessment
🔬 Research Areas
- • Machine learning refinement
- • Biomarker correlation studies
- • International formulation tracking
- • Novel ingredient classification
- • Real-time reformulation updates
Key Technical Takeaways
- ✓Additive scoring system: Base score (1-5) plus cumulative penalties (0.1-2.0) for industrial ingredients.
- ✓270K ingredients classified: Normalized and categorized into 84 functional buckets for consistent scoring.
- ✓Evidence-based penalties: Each penalty reflects scientific literature on processing impacts.
- ✓Transparent methodology: Every score component is traceable and explainable.
- ✓Continuous improvement: Regular updates based on new research and user feedback.
Apply This Knowledge
Disclaimer: All tools and data visualizations are provided for educational and informational purposes only. They are not intended as health, medical, or dietary advice. Always consult healthcare professionals for personalized dietary guidance.