Two food scanners can show different results for the same packet because they use different scoring methods, different product information or different personal settings. Sometimes there is a data error. Sometimes both apps are applying their own rules correctly.
The useful next step is to open the explanation. A number out of 100 looks familiar, but that shared scale does not mean the apps measure the same things.
We make Cull, so we have an interest in this question. The checks below apply to our results too.
Start with what each score measures
A nutrition rating, a processing classification and a combined app score are different tools. Before comparing their verdicts, look for the published method: which factors are included, how are they combined, and can any single factor limit the result?
For example, Yuka's published food method gives nutritional quality a 60% weighting, additives 30% and organic certification 10%. It also describes a score cap when an additive meets its own high-risk category. Those are Yuka's scoring rules, not percentages of a product's effect on your health (Yuka: how food products are rated).
Cull presents nutrition, additives and NOVA processing information, with explanations of its rating. A method that includes processing is asking a different question from one that rewards organic certification. Comparing the overall numbers alone hides that distinction. How Cull works.
A processing group is not a nutrition score
NOVA groups foods by the extent and purpose of processing. It does not calculate their nutrient balance. The UK's Scientific Advisory Committee on Nutrition has also identified difficulties in applying the classification consistently. Its April 2025 evidence update retains concerns about the limitations of the research (SACN: processed foods evidence update).
This means a processing classification cannot answer every shopping question. You may still want to know how much fibre, salt or saturated fat the product contains. Equally, a favourable nutrition rating does not tell you which manufacturing processes were used.
If two apps disagree on NOVA, look for the ingredient or process each identifies. A vague label may leave genuine uncertainty. Our NOVA guide explains the groups and their limits.
Check that both results describe your packet
Before debating algorithms, compare the basics: barcode, brand, product name and flavour. If an app displays a photograph, check that too. A similar name or familiar pack design is not enough to establish that two entries represent the same product.
Then compare the ingredient list and nutrition values with the packet in your hand. If one entry has different sugar or salt values, you have found a possible reason for the disagreement before reaching the scoring formula.
When something does not match, treat that result as unresolved. Use the app's correction or support route to report the mismatch, with the barcode and clear label photographs where requested. Keep the distinction between “different method” and “different information”: they need different explanations.
Make sure the quantities match
One result may describe 100g while another shows a serving. Preparation can matter too: dry food and food prepared with added liquid are different comparison bases. Check the headings before treating different figures as an error (DHSC: nutrition labelling guidance).
For a worked example, see per 100g versus per portion. Matching the quantity helps you decide whether the apps actually disagree about the product's nutrition.
Read what an additive flag means
An app's additive category is part of that app's assessment. Read its explanation and sources before interpreting a coloured dot as a claim that the food is dangerous.
The Food Standards Agency explains that approval of an additive includes assessment of safety for its intended use. The presence of an E-number alone therefore does not establish harm (FSA: food additives).
That does not make every question about additives settled. It means the relevant questions are specific: which substance, which use, what evidence, and what uncertainty? A count of unfamiliar ingredient names cannot supply those answers.
Check personal settings, then make your decision
Cull can adjust a product's score using your dietary profile. If you compare a personalised result with a standard result, the settings may explain the difference. Check the selected profile and the reason shown for any adjustment. Dietary profiles are included in Cull's free features (Cull features).
Keep allergy suitability separate from an overall score. Check the ingredients and precautionary warnings on the pack. In particular, a vegan label does not guarantee the absence of milk or egg cross-contamination (FSA: allergy information for consumers).
Once the product, quantities, method and settings match, you can judge the explanation. For your next shop, try one familiar barcode in Cull and open the rating breakdown. The useful result is understanding what is driving the number, so you can decide whether it matters for the food you are choosing.