Guides
Two pieces written to be useful to someone who never buys anything from us. Both are methods rather than opinions, and both work on any product in this category.
Two longer guides. The first is a ninety-second method for reading any liquid supplement label. The second is a three-question method for comparing cognitive supplements.
Neither argues for this product. Applied to this label, the first one produces the best-case page.
The two guides
Reading a liquid supplement label
Where the volume, the serving and the servings-per-container have to agree, what a dropperful is, and the four checks that take ninety seconds.
Choosing a memory supplement
A method for comparing products that does not rely on anything a product page tells you, and the three questions that eliminate most of the category.
Why these two
Almost every question a buyer asks about a supplement reduces to one of two things: what is actually in this bottle, and how does it compare with the alternatives. The first is a label-reading skill and the second is a comparison method.
Neither guide argues for this product. The first one, applied to this label, produces the best-case page. That is the test of whether a method is honest: it has to be allowed to reach an inconvenient answer.
The shorter pieces
Three shorter articles cover single topics in more depth than a product page allows.
The blog index →Proprietary blends explained →The grapefruit seed question →
See the Cognizil packs
Two, three or six bottles, priced plainly, with a 180-day window on the sales page kits. Read the best case first, then decide.
Shipping is calculated at checkout. This site publishes no figure it was not given.
What a method gets you that an opinion does not
A product page tells you what to think about one product. A method tells you how to think about all of them, including the one being sold on the page you are reading.
| Guide | The question it answers | How long it takes |
|---|---|---|
| Reading a liquid supplement label | What is physically in this bottle, and can the numbers on the label even close? | About ninety seconds per label |
| Choosing a memory supplement | Is this product comparable with anything, and does the evidence behind its ingredients survive the dose question? | About five minutes |
Scroll this table sideways →
Both are written so that they can be applied to a bottle in a shop without a phone. Neither requires trusting a website, including this one.
The test of an honest method
It has to be allowed to reach an inconvenient answer. Applied to this label, the first guide produces the best-case page, which is the least flattering document on this site. That is the point.
Where each method came from
Neither guide is original. Both assemble things that are already public into a sequence short enough to remember.
- The ceiling rule follows from 21 CFR 101.36(c), which requires descending order of predominance by weight inside a proprietary blend. The arithmetic is one line.
- The extract-ratio caution follows a 2022 review of extract declarations, which sets out why a native ratio needs a solvent, a plant part and a marker assay to be interpretable.
- The evidence hierarchy is the standard one used in any evidence review, compressed to six rows.
- The label arithmetic check — serving size times servings equals net contents — is required by the labelling rules themselves.
- The advice to establish a baseline before starting is from ordinary trial design, and it is the part most people skip.
Everything above is checkable, which is the property that matters. A method a reader cannot verify is an opinion wearing a numbered list.
Using the guides on a competing product
A retailer publishing a method that works against its own product invites an obvious question: what happens when you run it on the competition?
Usually something similar. Proprietary blends are the norm in this category, totals in the low hundreds of milligrams across a dozen names are common, and the ceiling rule bites the same way everywhere. A few products pass cleanly, and they tend to be the ones with three ingredients and a fully disclosed panel.
We are not going to name competitors, run their numbers, or publish a comparison table, because a retailer's assessment of a rival is worth nothing and everybody knows it. Run the method yourself. It takes ninety seconds and the answer belongs to you rather than to us.
The label method →The comparison method →Both, applied here →
Three shorter pieces
Each post takes a single question further than a product page allows.
What 15 milligrams looks like
Comparing a blend total with familiar objects. A 325 mg aspirin weighs twenty-one times the whole blend.
Extract ratios explained
What 20:1 states, what it leaves open, and why this site grants every ratio in full anyway.
The grapefruit seed story
Two published analyses of commercial grapefruit seed extract, and what they found instead of grapefruit.
A worked example, on this label
Rather than describe the method twice, here it is run against the Cognizil panel, with a stopwatch.
Seconds 0 to 5 — does the arithmetic close?
Serving size 1 ml, servings per container 30, net contents 30 ml. 1 × 30 = 30. It closes exactly, which is what you want.
Seconds 5 to 15 — what is quantified?
One dietary ingredient: chromium, 7 mcg, 20% DV. The DV for chromium is 35 mcg, and 7 is 20% of 35. That figure checks out.
Seconds 15 to 25 — find the blend total
Proprietary Blend, 15.47 mg, printed to the right of the name as the regulation requires.
Seconds 25 to 40 — count the names
Twenty-one. Divide: 15.47 ÷ 21 = 0.737 mg average if it were split evenly, which descending order says it is not.
Seconds 40 to 60 — check position one
Maca root extract. Ceiling 15.47 mg. Studied from 1.5 g a day. The entire blend is a hundredth of the smallest studied dose of its own first ingredient.
Seconds 60 to 75 — check the bottom
L-carnitine HCl, position 21, ceiling 0.737 mg, studied from 250 mg. The bottom ten positions are amino acids studied in grams.
Seconds 75 to 90 — read Other Ingredients
Water, natural flavors, potassium sorbate, stevia, xylitol, citrus extract. Note the xylitol if you have a dog.
Ninety seconds, no calculator beyond division, no product page, and you now know more about this formula than the front panel intended to tell you. The best-case page does the same thing for all twenty-one positions and adds the extract ratios on top.
What the guides deliberately leave out
Two large topics are absent from both guides, and the omissions are deliberate rather than accidental.
- Manufacturing quality
- Contamination, identity substitution and potency variance are real problems in this industry, and none of them is visible from a label. A guide cannot teach a reader to detect them; only a test document can, which is the subject of the testing page.
- Whether any of it works at all
- The guides answer whether a product can deliver a studied dose. Whether that dose does anything useful is a separate question, answered by the literature on each ingredient rather than by a label-reading method.
Keeping those two out is what allows the guides to be short enough to use. A method that tried to settle everything would settle nothing in ninety seconds.
If you only remember one thing
Find the blend total. Count the names. Divide. Compare the first position against the smallest dose in the research for that ingredient.
If the entire blend total is smaller than the smallest studied dose of its own first-named ingredient, nothing in that blend reaches a researched dose, and no amount of ingredient count, extract ratio, delivery format or marketing language changes it.
That test takes twenty seconds, it requires no trust in any website, and it settles a large fraction of this market. Everything else on this site is a longer version of it.
Who these are written for
Three kinds of reader arrive at a page like this, and the guides are built to serve all three without wasting anyone's time.
Someone holding this bottle
The worked example above is your ninety seconds. Then read the best-case page, which is the same method run to completion.
Someone comparing two products
Run the label method on both and compare the ceilings. The comparison method covers the part the labels cannot settle.
Someone who will never buy anything from us
Take the method and go. That is a legitimate use of these pages and it is the reason they are written as methods rather than as arguments.
The third reader is the one that keeps these guides honest. A guide written only for a buyer would quietly stop being a guide.
Corrections to the guides
A method with an error in it propagates that error into every label somebody runs it against, which makes these two pages the ones we would most want to be told about.
If a step does not work, a citation does not say what we say it says, or the arithmetic is wrong, write to [email protected] and name the step. Corrections are made on the page with the date.