How the DoseAudit score works: four scores, one verdict

DoseAudit scores every offer on four separate scales from 0 to 10: evidence, price, refund and transparency. The overall score is the plain average of the four, and that average sets the verdict: 7.0 or more is "Holds up", 4.5 to 6.9 is "Worth a look", and anything below 4.5 is "Doesn't add up". Nothing is weighted, and a commission never moves a number. The rest of this guide shows what each score looks at, with real reviews as worked examples, so you can check our math or repeat it on any product.

The score rates the offer: the label, the sales page, the checkout and the refund terms. We did not buy or use the products, and a score is not a medical opinion. It cannot tell you whether a supplement will do anything for you; ask a doctor or pharmacist about that.

What does the evidence score measure?

For a supplement, the evidence score asks one question: does the label match what the research tested? We copy each ingredient and its amount per daily serving, find the human studies for that ingredient, and compare the doses. We also open the studies the sales page cites and check that they tested the same ingredient, strain and dose.

Each ingredient row gets a status label (in studied range, below studied dose, can't verify, or mismatch), and those labels feed the score. The score is set by judgment against a written guide, not by a formula that counts labels, so two products with the same number of "can't verify" rows can land a little apart. The pattern is easy to see in real reviews:

• The GlucoBerry review found a stated amount for all four ingredients, and its 180 mg maqui berry dose matches the amount used in a small 2016 trial. Evidence: 5.0.
• The GlycoMute review found six named ingredients with no dose for any of them, so none can be compared with a trial. Evidence: 2.0.
• The Blood Sugar Blaster review found doses printed for most ingredients, but cinnamon at 50 mg against about 3 g in the trial the page cites. Evidence: 3.5.

The takeaway is that a printed dose earns more than a hidden one, and a printed dose that matches a trial earns more than a printed dose far below it. A high evidence score still describes the label, not the result you would get.

How does the price score work?

The price score uses the real cost per month, not the headline price. We follow our link to the order form and record every bundle, the shipping, any auto-ship box and any extra offer shown after payment. Then we divide by the days each bundle lasts at the label serving.

Bundle pressure matters here. Many offers show a low per-bottle price that only applies to a six-bottle order, and the one-bottle or two-bottle price is far higher. The same bottle can cost very different amounts depending on how many you commit to. As a rough guide, we treat about $30 a month as inexpensive and about $70 a month as expensive for a supplement.

Two examples from the blood sugar category show the effect:

• GlucoBerry has no single-bottle option. Two bottles cost $138 (about $69 a month) and six bottles cost $234 (about $39 a month). Price score: 5.5.
• The Berberine B1G2 review prices three bottles at $49, which works out to roughly $16 to $20 a month once shipping is counted. Price score: 6.5.

The price score says nothing about quality. It only answers "what will this really cost me each month, and how hard does the page push me toward the biggest bundle?"

How is the refund score calculated?

The refund score looks at three things: the length of the window, who handles the refund, and the conditions attached. The order form is the contract, so when the sales page and the checkout state different windows, we score the checkout. Most of the products we review are sold through ClickBank, which is the retailer of record, so refunds usually go through ClickBank support; our ClickBank refund guide walks through the steps.

Across the 17 blood sugar supplements we reviewed, longer windows scored higher:

• 60-day windows scored 5.5 to 7.5.
• 90-day windows scored 7.0 to 8.0. GlycoMute got 7.5, although its guarantee text says a refund is issued minus shipping and handling.
• 180-day windows scored 8.0 to 8.5, including GlucoBerry (8.0) and Gluco Extend (8.5).

Conditions explain the spread inside each band. Berberine B1G2 sits at 6.0 for a 60-day window because its sales page repeats a 180-day guarantee while the ClickBank order form we reached set the window at 60 days. We scored the 60 days and flagged the gap. A refund promise that the checkout does not back up is worth less than a shorter one that it does.

What does the transparency score cover?

Transparency asks whether you can find out who is selling the product and whether the page tells the truth about itself. We look for a named company and named people, a full label rather than generic bottle renders, and ingredient lists that match between the page and the label. We also look for the tactics that signal a page built to push rather than inform: countdown timers, "limited stock" claims with no supply data, unnamed doctors, dramatized testimonials and customer counts with no source.

Two blood sugar reviews sit at opposite ends. GlucoBerry scored 6.5, the highest transparency score among the blood sugar supplements, because the label is specific. The Insufend review scored 3.5: it prints real doses for vitamin D3, vitamin B1 and chromium, but seven other ingredients share one 525 mg blend, and it cites a customer count with no source.

The FTC says claims about the health benefits of supplements need substantiation in the form of "competent and reliable scientific evidence". We use that standard as a reference point when we read a sales page, although we are a review site, not a regulator, and we do not decide whether a claim is legal.

How do four scores become one verdict?

We add the four scores, divide by four and round to one decimal. The verdict follows from that average alone. Here are two complete examples:

Offer	Evidence	Price	Refund	Transparency	Average	Verdict
GlucoBerry	5.0	5.5	8.0	6.5	6.3	Worth a look
GlycoMute	2.0	5.0	7.5	2.5	4.3	Doesn't add up

The table shows why we publish all four numbers. GlycoMute's refund score of 7.5 is higher than GlucoBerry's evidence score, but its weak evidence and transparency pull the average under 4.5. A generous refund window does not rescue a label that cannot be checked.

Across all 518 reviews on the site, 3 reached "Holds up", 349 landed on "Worth a look" and 166 on "Doesn't add up". Averaging puts most offers in the middle, and the top band is hard to reach because a product needs solid marks on all four scales. In the blood sugar category, no review reached 7.0, and the highest score among supplements was 6.3.

What do the status labels and the check date tell you?

Each ingredient row in a review carries one of four labels. "In studied range" means the label amount sits inside the range used in human trials. "Below studied dose" means the amount is printed but is smaller than what the trials used. "Can't verify" means the amount, extract or strain is missing, so there is nothing to compare. "Mismatch" means the page, the label and the research disagree, for example a cited trial that tested a different strain. The ProDentim review shows the strain case.

The date matters as well. Prices, bundles and refund terms change, so every review states the day we checked, and the audit describes the offer as it looked on that day. Most of our current blood sugar reviews were checked on September 28, 2026. If you read a review months later, treat the price and refund window as a starting point and confirm them on the order form.

How can you repeat this check yourself?

You do not need our site to apply the method. Use this list on any supplement offer:

1. Find the Supplement Facts panel. If the page shows only bottle images, note that.
2. Check that every ingredient has an amount per serving, or a blend with a total weight.
3. Look up the amount used in human trials, using sources such as NIH pages and PubMed.
4. Open any study the page cites and confirm it tested the same ingredient and dose.
5. Work out the cost per month at the quantity you would really buy, shipping included.
6. Read the refund window on the order form, not only on the sales page.
7. Note the red flags: unnamed doctors, countdown timers, unsourced customer counts.

Scoring each of the four areas yourself and averaging them will show you quickly whether an offer is strong in one area and weak in the rest. Our how we review page explains the same four steps in short form, and the blood sugar category lists every review where you can see the scores applied.

FAQ

Does a high DoseAudit score mean a supplement works?

No. The score rates the offer: label doses compared with research, real cost, refund terms and seller openness. Even the best-scoring offers have not proven that the finished product does anything for you. Talk to a doctor or pharmacist before taking a supplement, especially if you use medication.

Do affiliate commissions change the scores?

No. Some links on the site are affiliate links, and we may earn a commission if you buy, but every product goes through the same four checks. Products that fail are published with their verdict, and our affiliate disclosure explains the funding.

Why are most scores in the middle?

The average pulls toward the middle. A strong refund window lifts a score, but hidden doses and thin transparency pull it back down. Reaching 7.0 needs good marks on all four scales at once, which only 3 of 518 reviews have done.

Can a score change after publication?

Yes. When we find a new price, bundle or refund term, we update the review and its check date. The score is recalculated from the four updated marks.