What to look for in an ingredient scanner app

Eight questions to ask any scanner, each with a test you can run standing in the aisle. We make one of these apps, so the list is written to be checked against ours too.

By Alex Vasile, Founder, LuminellePublished

An ingredient scanner app should show the regulator entry behind every flag, know whether a product is rinsed off or left on, be honest about the concentrations it cannot see, have a photo fallback when the barcode misses, and say how it is paid for. Below are eight questions to ask any of them, each with a test you can run in the aisle.

We make one of these apps. That is a reason to read this list sceptically, and it is also why the tests are written to be run rather than believed: each takes about thirty seconds with any scanner installed, including ours. The ranking is a separate page; this one is the criteria.

Does it show the source behind a flag?

A flag is a claim, and the question is whose. Some apps rate an ingredient on their own scale, out of 10 or out of 20, and the number is the whole argument. Others point at a hazard database. A few link to the regulator’s own entry: the line in the EU register, the California bill, the Proposition 65 listing. The difference matters the moment you disagree with a flag, because a citation can be checked and a rating cannot.

The test

Scan anything, tap the first flagged ingredient and keep tapping. If you land on a primary source, a regulation or an assessment, the app is showing its work. If you land on a paragraph the app wrote, you are being asked to trust it. EWG Skin Deep lists the studies behind each hazard score; Luminelle links each flag to the register entry.

Does it know what kind of product it is looking at?

The same preservative is permitted at different levels in a face cream and a shampoo, and an entry written for hair dye says nothing about a lip balm. EU restrictions are frequently specific to rinse-off, leave-on, oral, nail, eye or hair use. An app that does not know what it is scanning has to apply the strictest reading everywhere, which is how a cleanser ends up flagged for something that would only matter if it stayed on the skin.

The test

Scan a cleanser and a moisturiser that share a preservative; phenoxyethanol is in most of both. Compare the reason each app gives. If the wording is identical for both products, the app is not reading the product type.

What does it do about a dose it cannot see?

Above 1%, an ingredient list is in descending order of weight; below 1%, any order is allowed; nowhere is there a percentage. Every scanner is working blind on concentration. The honest ones say so, in the reason next to the flag, and score a concentration limit less than a ban. Some quietly score “permitted up to 1%” as if it were a problem. We measured what that did to our own scores, and it was most of the deduction.

The test

Find an ingredient the app flags with a limit; sunscreen filters and preservatives are the easy ones. Read the reason. Does it tell you the limit, admit it cannot see the level, and weight the flag accordingly? Or does a permitted ingredient carry the same colour as a banned one?

What happens when the barcode is not in the catalogue?

Catalogues are built from barcodes, and barcodes miss: imports, indie brands, salon lines, anything reformulated since the catalogue last looked. The question is what happens next. Some apps stop. Some let you type the list. Some read it from a photo, matching each INCI name to the register and scoring it the same way a barcode is scored.

The test

Pick the least famous product in your bathroom and scan it. If the barcode is unknown, look for a way to photograph the ingredient list. INCI Beauty is built around the list rather than the barcode; Luminelle has a photo mode, capped at five a day on the free tier.

How much of your bathroom does it recognise?

Coverage is where the older and bigger apps genuinely win, and it is worth saying plainly. Yuka resolves almost anything on a European or North American shelf; Think Dirty has a decade of North American barcodes. A newer app will miss more. How much that matters depends on the answer to the previous question, and on what you actually own.

The test

Scan five products from your own shelf, not the store’s. Count the misses, then check whether each miss still had a route to a result.

Can it be told what matters to you?

A general score is the same for everyone. What you care about is not: a fragrance sensitivity, an essential-oil reaction, a pregnancy. The useful version of personalisation is a section that highlights the ingredients matching your profile without changing the shared score, so your scan of a product is still comparable with anyone else’s.

The test

Look for a profile or preferences screen. Set a sensitivity, rescan a product you know contains it, and check two things: whether it is called out, and whether the overall score stayed the same.

How is it paid for?

Every app in this category is free to download, so the question is what pays for it: a subscription, a shop, affiliate links, brand partnerships, advertising, or your data. Independence is a feature when the app’s job is judging products, and it is not something a screenshot can show you.

The test

Read the store listing’s in-app purchases and privacy label, then the app’s own pricing page. Yuka states that it is independent and funded by its premium tier. EWG Skin Deep is a non-profit database. Luminelle is subscription-funded, with unlimited barcode scans free, no ads and no data sold. For any other app, read its own page rather than ours.

Does it explain, or only score?

A number is a verdict; an explanation is something you can learn from. The apps that explain tell you what an ingredient does in the formula, why a regulator cares, and what would have to be true for the flag to matter. INCI Beauty is strong on the first of those. The regulatory-status readings are strong on the second and third.

The test

Read one full ingredient page in each app, for the same ingredient. Which one could you argue with?

The eight, on one card

Eight questions, a thirty-second test for each, and what a good answer looks like
QuestionThirty-second testA good answer
Source behind a flagTap a flagged ingredient and keep tappingYou land on a regulation or an assessment, not a paragraph
Product typeScan a cleanser and a moisturiser sharing a preservativeThe reasons differ, because the rules do
Dose it cannot seeRead the reason on a limited ingredientIt names the limit, admits it cannot see the level, scores it less than a ban
Barcode fallbackScan your least famous productA way to photograph or type the list, scored the same way
CoverageScan five products from your own shelfFew misses, and a route to a result for each miss
Your profileSet a sensitivity and rescanThe match is called out; the shared score does not change
FundingRead the in-app purchases and the privacy labelSays plainly what pays for it; no ads, no data sold
ExplanationRead one full ingredient pageYou could argue with it, because it shows its reasoning
Run the tests on the free tier of each app; none of them needs a subscription to answer.

Questions people ask

Is a higher score always a better product?
No. A score measures whatever the app measures, a risk category, a hazard rating or regulatory status, and never efficacy, texture or price. A product can be excellent and score badly on a hazard scale, or be fully compliant and dull. Use the score to find the flags, then read the flags.
Do I need a paid app to check ingredients?
No. Every app in this category scans for free. Paid tiers typically add history, unlimited scans of some kind, routines, or analysis over time. Decide after you have run the tests above on the free tier.
Can an app tell me how much of an ingredient is in a product?
No. Ingredient lists carry no percentages. Above 1% they are ordered by weight, below 1% they are not, and that order is all any app can read. An app that implies it knows a concentration is guessing; a good one tells you the legal limit and admits it cannot see the level.
Should I trust an app over my dermatologist?
No. A scanner reads regulation and databases; a clinician reads you. The two are useful together: bring the ingredient list and the flags to the appointment, rather than the score.
Which app should I actually pick?
It depends on which of the eight questions matters most to you. If it is coverage, Yuka or Think Dirty. If it is ingredient explanation, INCI Beauty. If it is a free, independent hazard database, EWG Skin Deep. If it is a flag you can trace to a regulator entry, with the product type and the 1% line taken into account, Luminelle. The full comparison is here.

Keep reading

Luminelle vs YukaLuminelle vs Think DirtyLuminelle vs EWG Skin DeepLuminelle vs INCI Beauty

Run the tests on ours.

Free to download, free to scan. Every flag links to the regulator entry behind it, and unlimited barcode scans cost nothing.

Download on the App StoreGet it on Google Play

Out now on iPhone and Android.

What to look for in an ingredient scanner app | Luminelle