Methodology
How we evaluate, and how rankings will work
This page is the contract. Nutrition guides already follow it. Supplement Best of lists will not launch until the product half can be executed without pretending.
01
Conflicts and commercial relationships
Nutritionist Guide is published by OCN LLC, which also owns Rephora Labs. We sell reviews: a brand can pay for coverage. Money buys a review, not a rank. A paid product with a worse score cannot outrank an unpaid product with a better one. Rephora Labs products are eligible for Best of lists and are labeled at the mention. Brands do not see copy before publication.
02
Evidence hierarchy
Human outcomes outrank mechanisms. Systematic reviews and large trials outrank single papers. Consensus guidelines are starting points, not scripture. Marketing blogs are not sources.
03
Dates and updates
Every guide shows when it was published and when it last changed. A reversal in the evidence should produce a visible update, not a silent rewrite.
04
Who a recommendation is for
We say who a recommendation is for, and who should ignore it. Kidney disease, pregnancy, and disordered-eating histories are not footnotes we bury.
How we read a nutrition study
We ask what was measured (a biomarker, a hard outcome, or a questionnaire), who was enrolled, what the comparison diet was, and whether the result survived in more than one setting. A feeding trial that replaces butter with soybean oil is not interchangeable with a food-frequency questionnaire that asks people to remember cheese.
We do not “debunk” an entire nutrient because one cohort was null. We also do not crown a food because a cell study looked dramatic. When the literature is split, the guide should read split, not like a courtroom closing argument.
Sample citations on launch articles are labeled as examples. They show the kind of source we intend to use. They are not a claim that every bibliographic detail has been re-verified for this preview.
Future supplement rankings
The Best of section is a placeholder until we can score products against a protocol, not a vibe. When a ranking ships, each category page will publish its weights in advance.
Label accuracy
Does the serving contain the stated dose of the named ingredient? Hidden proprietary blends score poorly.
Contaminants and quality
When a category warrants it, we will look for third-party testing, heavy-metal risk, and manufacturing documentation, and say what we could not verify.
Effective dose
A product that requires six servings to match the trial dose is not equivalent to the study it cites on the label.
Safety margin
Upper limits, stacking risk, and populations who should not use the product are part of the score, not a disclaimer under the fold.
Value
Cost per effective dose, after you discard the fairy dust. Cheap and empty loses. Expensive and accurate does not automatically win.
How we test
The bench comes after the protocol
What we will not score
- Influencer popularity or unboxing aesthetics.
- A brand’s “clinical strength” language without a matching trial.
- Proprietary blends that hide dose.
- Products whose only evidence is a testimonial reel.