How the Milo Score Works: Our Food Score App Methodology

The Milo Score is a number from 0 to 100 that sums up how a packaged food stacks up on nutrition, additives and processing. It follows the methodology that Yuka has published, which in turn builds on Nutri-Score for nutrition and the NOVA classification for processing. Nutrition per 100 g carries the most weight, additives come next, and organic certification adds a small bonus.

A food score app is only useful if you can see how the number was reached, so this page sets out every part of the calculation and walks through a fictional muesli bar step by step.

The four bands

Every scored product lands in one of four bands, each with its own Milo monkey face so a child can read the result at a glance.

Band	Score range	What it means
Excellent	75 to 100	Good nutrition, no concerning additives, little processing
Good	50 to 74	A sound everyday choice with minor drawbacks
Poor	25 to 49	Notable issues with sugar, salt, fat, additives or processing
Bad	0 to 24	Several serious issues, or one additive with strong evidence of harm

The bands are a summary, not a verdict on you or your child. Plenty of Poor and Bad products are fine as an occasional treat.

Part one: nutrition per 100 g

The largest share of the score comes from nutrition, assessed the way Nutri-Score does it. Nutri-Score is the front-of-pack scheme developed by Santé publique France (santepubliquefrance.fr) and adopted across several European countries. It scores a food on a fixed set of nutrients:

Negative points for energy (kilojoules or calories), sugars, saturated fat and salt.
Positive points for fibre, protein, and the proportion of fruit, vegetables, legumes and nuts.

Each nutrient is banded into points, the positives are set against the negatives, and the result becomes a share of the Milo Score.

Why per 100 g and not per serving

Everything in the nutrition part is calculated per 100 g (or per 100 ml for drinks). This is the single most important design choice in the score.

Serving sizes are set by the manufacturer. A cereal maker can quote a 30 g serving to make the sugar look modest, while the bowl a child actually pours is closer to 45 g. Per 100 g removes that lever and puts every product on the same scale regardless of packet size. The trade-off is that something eaten in tiny amounts, such as a stock cube, can look harsher than its real-world impact. We accept that in return for a comparison that cannot be gamed.

Part two: additives

The second largest share of the score comes from additives. Each additive Milo recognises is placed in a risk tier based on published scientific assessments, with the evidence summarised in plain words in the app:

Strong evidence of harm. Milo caps the product’s score low, regardless of how good the nutrition is. A product cannot score above the Bad band with one of these present.
Some evidence of harm. Meaningful but less conclusive evidence. These pull the score down but do not cap it.
Limited or no concern. Little evidence of harm at the levels used in food, and little or no effect on the score.

Why one additive can drag a score down

Nutrition and additives are not simply averaged. If they were, a sugar-free sweet with an artificial colour linked to strong evidence of harm could still score well, because the nutrition part would mask the additive. Milo treats a strong-evidence additive as a ceiling instead, so a parent who wants to avoid that ingredient gets a clear signal rather than one buried under good fibre numbers.

The harms are named plainly. An artificial colour is called an artificial colour. Where the concern is about gut health, Milo says “disturbance of the gut microbiota” or “weakened gut barrier” and explains what that means. The evidence base draws on assessments from bodies such as the European Food Safety Authority (efsa.europa.eu), Food Standards Australia New Zealand (foodstandards.gov.au) and the US Food and Drug Administration (fda.gov). Our guide to food additives to avoid for kids lists the ones that come up most often in children’s foods.

Part three: NOVA processing group

NOVA, developed by researchers at the University of São Paulo, sorts foods by how much industrial processing they have been through, from group 1 (unprocessed or minimally processed, such as oats or plain milk) to group 4 (ultra-processed, built from industrial ingredients and additives you would not find in a home kitchen).

Milo shows a product’s NOVA group and uses it in the score. A group 4 product is marked down; a group 1 or 2 product is not. Because additives are themselves a marker of ultra-processing, the NOVA contribution is kept modest to avoid double counting. We explain the four groups in ultra-processed food for kids: NOVA explained.

Part four: the organic bonus

Products with a recognised organic certification receive a small bonus. Organic sugar is still sugar, so the bonus can nudge a borderline product across a band but cannot rescue poor nutrition. Milo applies it because Yuka’s published methodology does, and because it reflects lower pesticide exposure.

Putting it together: a fictional muesli bar

Here is a made-up product, Sunny Oat Choc Chip Bar. The numbers are illustrative.

Label, per 100 g: energy 1,850 kJ, sugars 28 g, saturated fat 7 g, salt 0.4 g, fibre 6 g, protein 8 g. Fruit and nut content about 5%. Ingredients include oats, glucose syrup, sugar, chocolate chips, vegetable oil, honey, an emulsifier and a natural flavour. Not organic.

Nutrition. Sugars at 28 g per 100 g and the energy density score heavily against the bar. Saturated fat is moderate, salt is low. Fibre and protein earn positive points back because the bar is oat-based. Net result: a middling-to-weak nutrition contribution, landing the bar around the low forties before anything else is considered.
Additives. The emulsifier is in the “some evidence of harm” tier for its effect on the gut microbiota, so it pulls the score down a further step. No additive with strong evidence of harm is present, so no cap applies.
NOVA. Glucose syrup, the emulsifier and the added flavour put the bar in NOVA group 4, which takes a few more points.
Organic. No certification, no bonus.

Result: 34 out of 100, Poor. Milo’s take would read something like: “High in sugar for a bar, with three different sugars in the list and an emulsifier that some studies link to gut irritation. The oats give it useful fibre.” The swap screen would then offer a higher-scoring bar from the same category.

Change the recipe and the score follows. Remove the emulsifier and replace the glucose syrup with dates and the bar rises into Good. Add an artificial colour with strong evidence of harm and it drops into Bad no matter what else changes.

Why some products show no score

Milo’s product data comes from the Open Food Facts database plus products that Milo users add by photographing the label. Some entries are incomplete: the nutrition panel or the ingredient list may be missing.

When there is not enough data to score a product honestly, Milo shows no score rather than a guessed one. A wrong number is worse than no number. If you hit an unscored product, photograph the label in the app and it will be added, so the next parent who scans it gets a proper result.

Allergen alerts sit outside the score

Allergens do not affect the Milo Score, and that is deliberate: whether a product contains dairy is not a matter of quality but of whether your child can eat it. You choose your allergens once (Dairy, Eggs, Fish, Soy, Peanuts, All nuts, Sesame, Sulfites, Gluten, Corn) and every scan shows an “Allergens detected” badge if the ingredients or “may contain” traces match. There is one allergen list per account.

The badge is an aid to reading the label, not a substitute for it. For a diagnosed allergy, always read the physical pack and follow your GP’s or paediatrician’s advice.

How swaps are chosen

For any product in the Poor or Bad band, Milo looks for a higher-scoring product in the same category and shows the two side by side. The swap has to be a realistic substitute: a cereal for a cereal, a corn chip for a corn chip. Two real examples from the app: Lucky Charms at 9 out of 100 with Three Wishes cereal at 90 as the suggested swap, and Doritos Nacho Cheese at 27 with Frontera Lime & Sea Salt chips at 80. Our healthy lunchbox snack swaps guide applies the same idea to a week of school snacks.

The independence pledge

A food score is only worth trusting if nobody can buy it. Milo carries no advertising, no affiliate links and no brand partnerships. Brands cannot pay to change a score, to be featured as a swap, or to have a product removed. The methodology on this page is the whole methodology. If you believe a product’s data is wrong, you can report it from within the app and we correct the entry. That is the only route by which a score changes.

How Milo helps

Milo turns everything above into a three-second scan: the score, the band face, Milo’s take in plain words, any allergen badge, and a better swap if one exists. Milo is free to download on the App Store, with Android coming soon. For the score at work in the cereal aisle, see hidden sugar in kids’ breakfast cereal.

Common questions

Is the Milo Score the same as Nutri-Score?

No. Nutri-Score covers nutrition only. The Milo Score uses a Nutri-Score-style nutrition assessment as its largest component, then adds additives, NOVA processing and an organic bonus, following Yuka’s published approach.

Does a Bad score mean a food is dangerous?

No. It means the product scores poorly on nutrition, additives or processing, or contains an additive with strong evidence of harm. Occasional treats are part of a normal childhood; the score is there to help with everyday choices, not to police them.

Why did two similar products get very different scores?

Usually a large difference in sugar or salt per 100 g, an additive in a higher risk tier, or one product being ultra-processed while the other is not. Milo’s take on each product names the reason.

Can I see the score for a product that is not in the database?

Not until it is added. Photograph the label in the app and the product is added for you and everyone else.