Why Do Ingredient-Scanning Apps Disagree With Each Other?

Why Do Ingredient-Scanning Apps Disagree With Each Other?

Scan the same bottle of sunscreen in Yuka, Think Dirty, and EWG Skin Deep and you can get three different scores: one green, one yellow, one red. The apps are not broken. They are measuring different things and weighting them differently, so a disagreement is expected, not a bug.

What Each App Is Actually Scoring

Yuka scores food and cosmetics separately. For food, it weighs nutritional quality, additives, and whether a product is organic. For cosmetics, it scores ingredient-level health risk based on hazard databases, with heavier penalties for endocrine disruptors and known allergens. A single 0-100 number comes out the other side, but the inputs behind that number change depending on the product category.

Think Dirty is cosmetics and personal care only. Its "Dirty Meter" score leans on a proprietary weighting of ingredient hazard data, with particular emphasis on ingredients linked to allergies and endocrine disruption. Because the full weighting formula is not published, two products with a similar ingredient list can land on different sides of the scale for reasons that are hard to audit from the outside.

EWG Skin Deep scores individual ingredients on a 1-10 hazard scale using published hazard and toxicity data, then rolls those scores up to a product-level score. EWG's method is largely ingredient by ingredient. It does not adjust much for how ingredients interact in a finished formulation, for concentration in the product, or for how the product is actually used.

Why the Same Product Gets Different Verdicts

Three things drive the disagreement.

Different inputs

Yuka pulls from nutrition and hazard databases. Think Dirty leans on a proprietary hazard weighting. EWG pulls from its own ingredient hazard database. None of the three use the same source data, so a difference in verdict often starts with a difference in what each app is even looking at.

Ingredient-level vs formulation-level scoring

EWG mostly scores ingredients in isolation. A preservative that reads as higher-risk on its own can behave very differently at a fraction of a percent in a finished product than the ingredient-level hazard score implies. Apps that do not adjust for concentration or formulation context will flag that ingredient the same way regardless of how much is actually in the bottle.

What counts as a red flag

Fragrance is a good example. Some apps treat "fragrance" or "parfum" as an automatic point deduction because it can hide undisclosed chemicals. Others weigh it less heavily unless a product also lists known allergens. Neither approach is wrong. They are different rules for how much unknown information should count against a product.

How Label Lookout's Approach Differs

We do not assign a single numeric score. Instead we evaluate against a published, category-specific rubric: what to seek, what disqualifies a product outright, and a list of harmful ingredients we check every candidate against. The full method is in How I Evaluate Every Product on This Site.

This does not make our method more correct than a scanning app. It makes it different, and transparent about where it draws its lines. For the fast version, The 5-Second Label Check That Changes Everything covers what we look for first when we pick up a product in a store.

We also do not treat every hazard database entry as automatically disqualifying. What EWG Verified Actually Means (and When It Does Not Matter) covers why a single certification, or a single low ingredient score, is not the whole story.

What To Do When the Apps Disagree

• Read the actual ingredient list yourself. Every scanning app is a shortcut for this step, not a replacement for it.
• Check what the app is actually scoring: food nutrition, cosmetic hazard, or both. A green score on a food app says nothing about a cosmetic hazard concern.
• Look for the reason behind a low score, not just the number. A single flagged ingredient at a trace amount is a different situation than a product built around several disqualifying ingredients.
• Treat conflicting scores as a reason to look closer, not as proof that one app is right and the other is wrong.

The Bottom Line

Yuka, Think Dirty, and EWG Skin Deep disagree because they measure different inputs, weight them differently, and draw the line between fine and flagged in different places. None of them is lying to you. Use any of them as a starting point, then check the ingredient list against what actually matters for the category you are shopping in.