How We Review
Here’s the process behind every review on this site, written out so you can judge our verdicts instead of trusting them.
What we check
1. Ingredient disclosure
Does the manufacturer publish a full ingredient list with quantities? Formulas that hide everything behind a “proprietary blend” with no dosing information get marked down under Transparency. You can’t evaluate a formula you can’t see.
2. Price and checkout terms
We open the manufacturer’s own order page and note:
- the current listed price, and any multi-bottle discount tiers
- whether it’s a one-time order or turns into a subscription
- the stated refund window, and whether the terms match the headline claim
The date of that check is printed on the review as Facts checked. Prices change, so we re-verify on a rolling basis and move the date when we do.
3. Claims versus published evidence
For each headline claim we ask one question: would a reasonable person reading the published literature reach this conclusion? Where a manufacturer states something the evidence doesn’t support, we say so. Where the evidence is simply thin — which is normal for newer ingredients — we say that too.
4. Customer-reported experience
We gather publicly visible customer feedback and set aside the entries that just repeat the marketing copy. Then we report the positive themes and the negative ones. A product with no complaints anywhere gets treated with suspicion, not enthusiasm.
What we score, and how
Reviews score out of 5 across five dimensions:
| Dimension | What it measures |
|---|---|
| Ingredient transparency | Is the full formula and dosing published? |
| Value for money | Price per serving against comparable products |
| Purchase terms | Refund window, subscription clarity, checkout honesty |
| Claim accuracy | Do the marketing claims match the evidence? |
| Customer-reported results | Volume and consistency of real-user feedback |
The overall score reflects those five together. It isn’t a weighted average tuned so everything lands at four stars. A product that fails on transparency scores badly no matter how popular it is.
How we grade evidence
Product scores measure a bottle. They do not measure whether the ingredient works. Those are separate questions, so we grade them separately, on the evidence tables inside our ingredient articles.
Every row in those tables carries one of three grades. The grade describes how much confidence the published data supports, not how impressive the number looks.
| Grade | What it requires | What it does not mean |
|---|---|---|
| A | Multiple randomised trials, consistent direction of effect, population close to the people who would use it | That the effect is large |
| B | One large well-conducted trial, or several trials with material heterogeneity between them | That the trials agree |
| C | Small samples, mixed formulations, populations that differ from the reader, or a pooled estimate with an interval near zero | That the ingredient is useless |
Three rules govern how a grade is assigned, and they are the reason some familiar claims score lower than their marketing suggests.
An interval crossing zero cannot earn a grade above C. If the 95% confidence interval includes no effect, the trial did not establish one, regardless of where the point estimate sits. “Showed a trend toward improvement” describes exactly this case.
Population mismatch caps the grade at C. Most glucose trials enrol people who already have type 2 diabetes. A result in that group does not transfer to someone acting preventively, and we do not grade it as if it did.
Strain, form and dose must be named. For probiotics, an effect belongs to a strain code, not a species. For botanicals, an extract standardised to a marker is not interchangeable with raw powder. If a trial does not specify what it used, the row cannot be applied to a product and is capped at C.
Why we publish this
Several supplement databases assign grades without showing the rule behind them. A grade you cannot audit is a verdict, not a measurement. Writing the criteria down means a reader can disagree with us on the merits — take a row we graded B, apply the standard, and argue it should be C. That argument is only possible if the standard is visible, and it is the reason we print it here rather than keeping it internal.
Things we will never do
- Claim a supplement cures, treats, prevents or heals a disease. They aren’t medicines. If a manufacturer hints otherwise, we flag it as a red flag in the review.
- Invent test data. We don’t run clinical trials. When we reference research, we link to it.
- Publish a score we can’t defend. Every score breaks into the five dimensions above, so you can see exactly where a product gained or lost points.
- Let commercial interests colour a verdict. A product we can’t recommend gets said plainly, and we publish those write-ups anyway.
When we can’t verify something
Some products are sold only through a video presentation with no readable ingredient documentation, or the seller’s site blocks automated price checks. When that happens we say on the page what we could and couldn’t confirm. We don’t fill the gap with guesses.
Health disclaimer
Nothing on this site is medical advice. These products are not intended to diagnose, treat, cure or prevent any disease. If you’re pregnant, nursing, taking prescription medication, or managing a medical condition, speak to a qualified clinician before starting any supplement.
Corrections
Spotted an error, or a price that’s moved? Email [email protected] with the page URL. We fix factual errors and note substantive updates on the page itself.