Our methodology

How we review.

The full HEME Review Framework, published in plain English. The six criteria, the scoring rubric, the editorial firewall and the rules every HEME editor follows.

Last updated 1 May 2026 · Reviewed quarterly by the editorial board · Version 4.2

Why we publish this at all.

Most product-review websites either don't publish a methodology, or publish a one-paragraph version that says "our experts review products and recommend the best." That sentence is doing a lot of work. If you can't see how the sausage is made, you can't trust the sausage.

The HEME Review Framework is published in full so readers, brands and regulators can see exactly how a product earns its score. There are no hidden criteria, no proprietary algorithms, and no "secret testing partners." If we change the framework, we publish the new version and we re-score affected reviews.

This document is the contract between HEME and its readers. If we break it, you should call us out — and we'll publish a correction.

The six criteria.

Every product reviewed on HEME is scored against the same six criteria. Each criterion is worth 20 points. The maximum possible HEME score is 100, calculated from a 120-point raw total normalised against a "claim integrity" penalty (see Scoring).

1

Bioavailability

Is the active ingredient in a form the body can actually absorb? For iron, this means form (bisglycinate, fumarate, heme polypeptide, etc.), supporting cofactors (vitamin C for non-heme iron) and the presence of absorption blockers in the same formula. For other categories, the equivalent question: is the form on the label a form with credible evidence of absorption?

0–5
Poor form / contains active blockers.
6–10
Standard form, no cofactor support.
11–15
Good form, sensible support matrix.
16–20
Best-evidence form with cofactor pairing.
2

Dose transparency

Does the label disclose elemental dose for each active, or only the salt weight? Are proprietary blends used to hide individual quantities? A product that hides its dose information cannot be assessed as well as one that publishes it openly — so we mark it down on this criterion accordingly.

0–5
Proprietary blends, no per-ingredient dose.
6–10
Partial disclosure (some doses).
11–15
Full disclosure of every active.
16–20
Full disclosure + clarification of elemental dose.
3

Third-party testing

Is the product independently tested for purity, heavy metals and label-claim accuracy? We weight named certifications (NSF, USP, Informed Choice, ISO 17025) more highly than self-reported testing. Public, batch-level certificates of analysis score higher than aggregate statements.

0–5
No third-party testing disclosed.
6–10
Self-reported, no certificate access.
11–15
Named certification, aggregate CoA.
16–20
Public batch-level certificates of analysis.
4

Tolerability

Real-world reports of side effects, taste, capsule size, smell, swallowability, and how the product fits into a daily routine for an actual woman trying to take it. We weight aggregated user reviews from multiple independent retail platforms over single-source testimonials. Where applicable, we conduct a small in-house tolerability panel.

0–5
Frequent reports of intolerance.
6–10
Mixed reports — some struggle.
11–15
Generally well-tolerated.
16–20
Consistently well-tolerated across users.
5

Price per serve

Cost of a standard serving at the time of review, including any subscription discount and shipping where relevant. We compare price-per-serve relative to category — we do not penalise products for being premium-priced if the formulation justifies it. We do penalise products that are expensive for what they actually contain.

0–5
Expensive and not justified by formulation.
6–10
Above category average.
11–15
Fairly priced for what's in the bottle.
16–20
Exceptional value for the formulation.
6

Claim integrity

Does the brand's marketing language sit honestly within what the formulation can actually do? We mark down products whose claims overreach — particularly any product using medical-claim language like cure, treat, fix, diagnose, or balance hormones without regulatory approval to do so.

0–5
Makes medical claims without basis.
6–10
Overclaims relative to evidence.
11–15
Marketing is mostly aligned with formulation.
16–20
Honest, evidence-aligned, no overreach.

How we calculate the HEME score.

Each criterion contributes up to 20 points. The raw total is normalised to a 100-point scale for the final HEME score, displayed on every review page. We've kept the formula simple — and published — so it can be checked.

The formula

(Sum of 6 criteria) × (120 / 100)

Each criterion scores 0–20. Maximum raw total is 120. We then normalise to a 100-point scale. Any product scoring below 60 will not appear on the front of a HEME shortlist — it goes into the "skip" tier with a published reason.

We do not weight criteria differently. A product that fails on dose transparency cannot compensate by being cheap. A product that fails on tolerability cannot compensate by having a great-looking certificate of analysis. A great HEME product is one that does the basics consistently — not one that excels in one criterion at the expense of others.

The editorial firewall.

HEME generates revenue from affiliate commissions, sponsored reviews, the HEME Reviewed badge programme, clinic partnerships, and newsletter sponsorship. None of these revenue streams change how a product is scored.

What we will do

  • Disclose every sponsored review at the top of the article and in the URL.
  • Apply the framework identically to sponsored and independent reviews.
  • Publish the score the framework produces, whatever it is.
  • Disclose affiliate relationships on every review page.

What we won't do

  • Edit a published review to hide a low score.
  • Remove a "cons" section at a brand's request.
  • Promise a positive review in exchange for payment.
  • Award the HEME Reviewed badge below criteria threshold.

If we break the firewall

If a HEME editor or executive violates the firewall, we will publish a correction at the top of the affected review, name what happened, and explain how we will prevent a repeat. Editorial integrity is not a marketing point — it is the asset.

How a review is actually made.

The same five-stage process applies to every review, sponsored or independent. The timeline from product selection to publication is usually four to six weeks.

From product to published review

01
Sourcing
Products selected based on category research, reader requests, and brand submissions.
02
Formulation review
Editorial team analyses the label, supporting documents and brand-provided certificates.
03
Framework scoring
Each criterion scored independently. Two editors score, then reconcile.
04
Drafting
Lead editor writes the review. The brand sees a draft for factual fact-checks only.
05
Publication
Review published with score, disclosures, and a public update log.

Who writes them.

HEME reviews are written by trained editorial staff and reviewed by qualified clinicians where the subject matter is clinical. Bylines on every review identify the lead editor, the clinical reviewer (where applicable), and the framework editor who calibrated the score.

How products get on our list.

We source products in three ways. Reader-requested reviews come from women emailing us about products they're seeing — these get priority for category round-ups. Category-driven reviews come from gaps in our coverage and trends in what readers are searching. Brand-submitted products go through the same scoring as any other product on our list — submission does not guarantee a review, and a review does not guarantee a positive outcome.

The For Brands page explains submissions in full. Read how to submit →

When we update reviews.

Reviews are not static. We re-evaluate when:

  • A product reformulates, changes manufacturer, or changes packaging information
  • New third-party testing data becomes available
  • A reader flags a meaningful new data point we missed
  • The category itself sees new evidence that affects how we'd score

Every review has a "last updated" date at the top, and we maintain an update log at the bottom of each page documenting material changes since first publication.

Affiliate disclosure.

HEME participates in affiliate programmes with several supplement retailers and brands. When a reader buys a product through a link on a HEME review, we may earn a commission. Affiliate links exist on most product cards across the site and are disclosed at the top of every review.

Affiliate commissions do not influence which products earn a place in our shortlists, the scores we publish, or the editorial language we use. Where two products score identically, we do not preference the higher-commission link. Where a low-scoring product has the highest commission, it is still listed as "skip."

The HEME Reviewed badge.

The HEME Reviewed badge is a separate, more rigorous assessment than a standard review. Brands apply for it; products either earn it or they don't. Badges are valid for 12 months and re-assessed annually. Products that fall below criteria threshold during the year lose the badge.

The badge cannot be purchased independently of meeting the criteria, and the application fee is non-refundable whether the product earns the badge or not. More on the badge programme →

Corrections policy.

If we get a fact wrong, we correct it as quickly as possible, transparently, and at the top of the affected page. Substantive corrections are logged in the page's update log with the date and the nature of the change. We will not silently edit reviews to walk back a published claim.

To request a correction, email corrections@heme.co. Every email is read by a senior editor.

HR
Hannah Reilly
HEME Reviews Editor — responsible for the framework
— signed and accountable.