The Evidence Standard
How we decide what we know.
HELM is a cosmetics company, not a research institution. What we can do is read the available evidence carefully, say where each claim comes from, and be clear about the difference between what has been studied and what a finished product does.
Evidence hierarchy
01
Regulatory and safety assessments
Ingredient safety reviews and regulatory determinations from official bodies. These carry the most weight for questions of permitted use, concentration limits, and labelling.
02
Systematic reviews and meta-analyses
Structured evaluations of the full body of published work on a question. Strongest for understanding overall direction across many studies.
03
Controlled human studies
Randomized and controlled studies on people. Useful and relatively direct, but often small, short, and run on a specific formula rather than a class of products.
04
Laboratory research
In vitro and ex vivo work on cells, tissue, or hair fibres. Valuable for mechanism, but does not establish what a finished cosmetic does on real skin under real use.
05
Supplier and technical documentation
Ingredient specifications, technical dossiers, and supplier studies. Necessary for formulation and quality work; read with an understanding of who produced it and why.
06
Formulator and expert consensus
Established practice within cosmetic formulation. Reliable for how things are done and why; not a substitute for study evidence about outcomes.
07
HELM internal testing
Our own stability, compatibility, and quality testing. It tells us the product performs and holds as specified — it is not published research.
Our principles
Ingredient research is not product evidence
A study on an ingredient at a given concentration, in a given base, over a given period, is evidence about that study. It does not automatically transfer to a finished product containing the same ingredient. We keep those two claims separate, always.
Concentration and formulation context matter
The same ingredient can behave differently at a different level, at a different pH, in a different base, or alongside different ingredients. Naming an ingredient tells you very little on its own.
Strength of evidence is stated, not implied
Where the evidence is strong, we say so. Where it is preliminary, limited, mechanistic, or mixed, we say that instead of dressing it up in confident language.
Language is chosen carefully
Cosmetics support the appearance, feel, and condition of skin and hair. We do not describe our products in medical or therapeutic terms, and we do not imply outcomes we cannot support.
Uncertainty is disclosed
Open questions are part of the picture. Saying what is not yet established is more useful to you than pretending the question is settled.
Positions are revisited
Evidence moves. When it does, we update the relevant pages and note that they changed rather than quietly leaving old claims in place.