Method and limits
How we score supplements.
Every De-Influenced report reads a label the same way: what is disclosed, whether the doses look meaningful, and whether the price is justified. This page explains those signals, where they come from, and what we deliberately do not claim.
Transparency is not the same as product quality.
A clear label can still contain weak doses. That is why the report shows dose context separately.
The starting point
Every report answers three questions.
Can you tell what is inside?
Amounts, forms and blends are read as label clarity signals. Hidden blend amounts reduce confidence.
Does the dose look meaningful?
Where a dose can be assessed, we say whether it appears meaningful, low or unclear for the role it plays.
Is the price justified?
Value is read alongside dose and clarity. Price alone does not make a product good or bad.
The verdict at the top of each report is our editorial reading of those three answers together. It weighs how many active ingredients appear meaningfully dosed, how clearly the label discloses what is inside, and how the price compares with similar products. It is never a medical recommendation.
The score
A 100-point reading of label clarity.
The Ingredient Transparency Score rewards labels that let you judge what you are buying. It is built from three pillars.
Dose Disclosure
Does the label state an individual amount for each active ingredient?
The biggest share of the score. Labels that disclose every active dose earn full points. Proprietary blends that show one total but hide the individual amounts lose points here, because you cannot tell what you are actually getting.
50
Ingredient Specificity
Does the label name the exact form of each ingredient?
Magnesium bisglycinate is not the same as magnesium oxide. Naming the specific form, salt or standardisation tells you what the ingredient can actually do, so labels that name forms score higher than labels that name only the ingredient.
35
Label Clarity
How much of the formula is genuine filler?
A formula padded with fillers is harder to read and usually working harder on the marketing than on the contents. Standard processing excipients are not penalised; genuine padding is.
15
A transparency score measures how clear the label is, not how good the product is. A perfectly transparent label can still describe a weakly dosed formula.
That is why every report shows dose context as its own signal, separate from the transparency score. Read the two together.
91–100
Excellent
Doses disclosed, forms named, minimal filler. You can see exactly what you are buying.
76–90
Good
Most doses and forms are disclosed. Small gaps, but the label is honest about its contents.
61–75
Moderate
Clear in places, vague in others. Some doses or forms are missing or hidden.
41–60
Below Average
Significant gaps. Key doses are undisclosed or buried in blends, so much of the formula cannot be judged.
0–40
Poor
The label tells you very little about what is inside. Treat any claims with caution.
Dose context
Present is not the same as meaningfully dosed.
Dose context is built from published human research, accepted nutrient reference values where relevant, product label data and our own review rules. It describes how a disclosed amount compares with studied ranges. It is not personal dosage advice.
Where a label does not give enough information to judge, we say Cannot Assess rather than guessing. Treat that as honesty, not a gap in the report. Some of the reference material behind our dose ranges is licensed, so we describe how we use it rather than republishing it.
Full Strength
The disclosed amount and form sit within the ranges used in human research for this ingredient.
Adequate
The amount looks meaningful, though it sits below the strongest studied ranges.
Underdosed
The amount is below the ranges normally used in research for the role this ingredient plays in the formula.
Sub-therapeutic
The amount is so far below studied ranges that it is unlikely to do what its presence on the label implies.
Excessive
The amount is above the usual studied ranges. More is not automatically better.
Cannot Assess
The label does not give enough information to judge, usually a missing amount or an unnamed form. We say so rather than guess.
Proprietary blends
When the label shows a total but hides the doses.
A proprietary blend shows one total amount for a group of ingredients but hides each individual dose. That makes it impossible to tell whether the headline ingredient is meaningfully dosed or only present in a token amount.
Blends are scored as undisclosed doses, so they lower the transparency score, and the ingredients inside them are usually marked Cannot Assess. That is not an accusation of bad intent. It simply reflects what the label lets a shopper verify.
Reading the research
A studied ingredient is not a proven product.
“Clinically studied” on a label usually refers to one ingredient, not the product in the tub. Before treating research as support for a claim, we ask six questions.
Was the study on the finished product, or just on one ingredient in it?
Was the studied dose the same as the dose in the product?
Was it the same form of the ingredient?
Was it tested in humans?
Was there a proper comparison group?
Does the benefit measured actually match the claim being made?
Value
Price only matters after context.
Value on a report is read in context, not from the price tag alone. We weigh the cost per serving against similar products, how many actives appear meaningfully dosed, and how transparent the label is. A low price is not a bargain if the useful doses are missing or hidden, and a premium price needs stronger label support to justify itself.
Independence
Funded by subscribers. Not by supplement brands.
De-Influenced is funded by subscribers, not supplement brands. There are no affiliate links, no sponsored placements and no ad revenue. Alternatives are chosen by the matching method, never by commission.
If a product we recommend is cheaper than the one you analysed, we earn nothing either way. That is the point.
The limits
What we deliberately do not claim.
A label has limits, and so do we
We analyse what a product discloses, not what happens in your body. We do not diagnose conditions, predict outcomes, classify products as medicines or replace your healthcare professional.
Reports age
Labels and prices change. A report reflects the product information available at the time it was run. If a product has been reformulated since, tell us and we will recheck it.
De-Influenced is an educational supplement label analysis tool. It does not provide medical advice, diagnose conditions, treat disease or classify products as medicines, and it does not tell you what supplements to take. Always speak to a qualified healthcare professional before changing your supplement routine.
Spot something wrong?
Think we have got something wrong? Email us the product URL and what looks off. A photo of the label helps us recheck quickly.
Anchored to official guidance
The public sources behind the method.
Our reading of supplement claims follows the rules that apply to supplements sold in the UK.
- CAP Code Section 15: Food, food supplements and associated health or nutrition claims
Advertising Standards Authority · The advertising rules that supplement claims in the UK must follow.
- Supplementary advice on health claims in ads for supplements
Advertising Standards Authority · How the ASA expects health claims for supplements to be substantiated.
- Great Britain nutrition and health claims (NHC) register
GOV.UK · The register of authorised and rejected health claims in Great Britain.
- Nutrition and health claims: guidance to compliance
GOV.UK · Official guidance on which nutrition and health claims may be made on foods.
- Food supplements: business guidance
Food Standards Agency · How food supplements are regulated and labelled in the UK.
- Borderline products: how to tell if your product is a medicine
MHRA via GOV.UK · Where the line sits between a food supplement and a medicine. De-Influenced does not classify products as medicines.
See it applied to a real label.
Paste a product link and read the report with this page beside it.
Analyse a product