Methodology

Dekost combines three independent data sources to give you a complete picture of what Dutch supermarket food actually costs β€” in euros and in nutritional quality.

Nutrition data β€” NEVO

Nutrient values come from the NEVO database (Nederlands Voedingsstoffenbestand), maintained by RIVM. NEVO is the Dutch national food composition table, compiled from laboratory analyses of representative food samples. It covers macronutrients, micronutrients, amino acids, and fatty acid profiles.

The foods you compare here are NEVO reference entries, not individual supermarket products. Where we hold a supermarket price, the product it was recorded for is linked to one of these foods, and that price is shown on the food.

Price data

Price coverage is thin, and we would rather say so than imply otherwise. Most foods in this database carry no price at all. Where a price exists it is expressed as €/kg so package sizes can be compared, and we record the base price rather than a promotional or loyalty-card price, because a discount you cannot rely on is not the cost of the food.

Two sources, and they are not equal. A small set of prices was entered by hand and is dated accordingly β€” old, but not wrong. Since September 2026 we also collect prices directly from Ekoplaza, whose robots.txt permits it and whose own structured product data we read rather than guess at.

We do not collect from Albert Heijn or Jumbo. AH's robots.txt disallows it, so we do not, and Jumbo's site blocks automated access. We identify ourselves in every request rather than pretending to be a browser, we obey robots.txt even when it is inconvenient, and we would rather have narrower data than data taken from someone who asked us not to. That means our price picture leans towards organic retail and is not representative of an average Dutch trolley.

Some products have no €/kg at all and correctly show none: an item sold per piece, or in millilitres without a known density, has no honest mass basis. We do not use affiliate links; all supermarket links go directly to the retailer.

Derived metrics

  • € per gram protein β€” price per kg divided by the grams of protein in a kg (protein per 100g Γ— 10). The primary efficiency metric.
  • € per 100 kcal β€” energy cost. Useful for calorie-dense foods like nuts and oils.
  • Protein density β€” grams of protein per 100 kcal. Measures protein efficiency relative to caloric load.
  • Complete protein ● β€” marked when a food meets all nine essential amino acid requirements at a 100g serving, based on WHO/FAO reference values cross-referenced with NEVO and USDA amino acid data.

Data sources

  • Β© Rijksinstituut voor Volksgezondheid en Milieu (RIVM), Nederlands Voedingsstoffenbestand NEVO Online versie 2025/9.0, 2025 β€” nevo-online.rivm.nl
  • Amino acid profiles β€” U.S. Department of Agriculture, Agricultural Research Service. FoodData Central SR Legacy, 2018. fdc.nal.usda.gov (US federal government work, public domain under 17 U.S.C. Β§ 105) + NEVO tryptophan data
  • Prices β€” a hand-entered seed set, plus direct collection from Ekoplaza (robots-permitted, from the retailer's own structured product data). Not Albert Heijn or Jumbo β€” see Price data above

How scores are calculated

Each food page shows several derived scores. Here is what each one means and how it is computed.

Nutri-Score

A five-letter label (A–E) computed live from NEVO nutrient data using the 2023 general-foods algorithm. It is not stored β€” recalculated fresh each time from the food's macros.

The algorithm works in two passes. First, it sums "negative" nutrient points (N): energy (kJ), saturated fat, total sugars, and sodium β€” each scored 0–10 on a stepped scale. Then it sums "positive" nutrient points (P): fruit/vegetable/nut/legume content, dietary fiber, and protein β€” each also 0–5. The final score is N βˆ’ P. Lower is better.

Score≀ βˆ’10 – 23 – 1011 – 18β‰₯ 19
GradeABCDE

One nuance: protein points are only added if N < 11 or the food is β‰₯ 80% fruit/veg/nut β€” otherwise the protein bonus is withheld to prevent very energy-dense, high-protein foods from scoring too favourably.

Reference: Nutri-Score 2023 Scientific Committee Technical Report, Serge Hercberg et al.

Protein Quality

Rates whether a food provides all 9 essential amino acids in sufficient amounts, using the WHO/FAO/UNU 2007 amino acid scoring pattern as the reference β€” expressed in milligrams per gram of protein. A food is rated complete when every essential amino acid meets or exceeds 100% of the reference value. The overall quality is constrained by the lowest-scoring amino acid, called the limiting amino acid.

This is conceptually similar to PDCAAS (Protein Digestibility-Corrected Amino Acid Score) but without digestibility correction. Data from NEVO 2025 (tryptophan only) and USDA FoodData Central SR Legacy (full profile, matched by name similarity).

Amino Acid Profile

A detailed radar chart showing all 9 essential amino acids scored against the WHO/FAO 2007 reference pattern per gram of protein. A score of 100% means the food exactly meets the reference; above 100% means it exceeds it. The red dot marks the limiting amino acid. The dashed ring marks the 100% reference level. Capped at 150% for visual clarity.

Data from USDA FoodData Central SR Legacy, matched to NEVO foods by name similarity. Where a USDA profile is used, the food page names the matched USDA food and shows its match percentage above the chart.

Caloric Ratio

Shows how the food's energy is distributed across the three macronutrients using their Atwater energy factors: carbohydrates at 4 kcal/g, fat at 9 kcal/g, and protein at 4 kcal/g. The ternary triangle chart places the food in a three-way macro spectrum β€” a food in the top corner is almost entirely carbohydrates; bottom-right is primarily protein.

Fullness Factor

Estimates how filling a food is per calorie, on a scale from 0.5 (barely filling) to 5.0 (very filling). Foods high in protein and dietary fiber tend to score higher; energy-dense foods with little fiber score lower. Formula by SELF Nutrition Data:

FF = 41.7 / CAL^0.7 + 0.05 Γ— PR + 0.000617 Γ— DFΒ³ βˆ’ 0.00000725 Γ— TFΒ³ + 0.617

Where CAL = kcal/100g, PR = protein (g/100g), DF = dietary fiber (g/100g), TF = total fat (g/100g). Clamped to [0.5, 5.0].

Nutritional Target Map

A radar chart showing how much of five daily macro targets a 100g portion covers: protein, fiber, fat, carbohydrates, and calories. Values are √-scaled (square root) so that small percentages remain visible and are not crushed by outliers. A food that broadly covers all five axes makes a balanced macro contribution.

Daily values used: protein 50g, fiber 28g, fat 78g, carbohydrates 275g, energy 2000 kcal (FDA/EFSA general reference values).

Nutrient Balance

A polar chart mapping a food against EU Nutrient Reference Values (NRVs) for vitamins and minerals per 100g. Each spoke represents one micronutrient; the length shows the percentage of the NRV that a 100g portion covers. A wide, round shape indicates a food that is rich across many vitamins and minerals. Nutrient data from NEVO 2025.

Vitamins and minerals tracked: vitamin A, D, E, C, B1, B2, B3 (niacin), B6, B12, folate, calcium, iron, zinc, magnesium, selenium, potassium, iodine.

Price & Protein Efficiency

Best price (€/kg) is the lowest price per kilogram we have actually observed for that food, at base price, excluding promotions and loyalty-card prices. It is the lowest of what we hold, not the lowest in the Netherlands β€” our sources are a hand-entered seed set and Ekoplaza, so read it as one observed price rather than a market survey. Most foods have none.

€ per gram protein = price (€/kg) Γ· (protein (g/100g) Γ— 10). This normalises the price to actual protein delivered.

Protein density = protein (g) Γ· energy (kcal) Γ— 100. Measures how much protein you get per calorie. The track shows this food versus the category average β€” a longer colored dot means above-average protein density for its class.

Percentile Scores

Each food is ranked against all others in the database for several metrics. Scores are expressed as percentiles (0–100) and "Top X%" labels, where Top 10% means the food ranks in the best 10% for that metric. Below the median the label reads "Bottom X%" instead, so the last food in a metric is "Bottom 1%" rather than "Top 100%".

  • Protein density score β€” percentile for g protein per 100 kcal across all foods. Score 73 = higher than 73% of all foods.
  • Price efficiency score β€” percentile for g protein per €/kg price. Only foods with price data are ranked.
  • Satiety score β€” percentile for Fullness Factor (SELF formula above).
  • Micronutrient density score β€” percentile for average % EU NRV covered per 100g across 17 vitamins and minerals.
  • Value score β€” average of protein density score + price efficiency score. Only computed for foods with price data.

Rankings are computed in your browser each time the dataset loads, across every food in it. Adding new foods changes all percentile scores.

License & source

Β© Dekost / Ryan Reddy. All rights reserved. Chart design, derived metrics, scoring formulas, data compilation, price tracking methodology, and all editorial text are original works of Dekost.

Application code is proprietary software. Personal and non-commercial use is permitted. Commercial use requires a license β€” get in touch.

Nutrient data: Β© Rijksinstituut voor Volksgezondheid en Milieu (RIVM), Nederlands Voedingsstoffenbestand NEVO Online versie 2025/9.0, 2025. All rights reserved by RIVM. Reproduced with attribution per RIVM's terms of use. USDA FoodData Central data is a US federal government work and is public domain (17 U.S.C. Β§ 105).

See the Licensing & Sources page for full details. If you find an error or want to contribute data, get in touch.