What a proprietary blend actually hides

Two words on a panel remove every individual amount from view, lawfully and in plain sight. Here is exactly what the rule permits, the fair argument for using it, and the arithmetic that decides whether a particular blend is worth your money.

The phrase is a legal permission, not a flourish

"Proprietary blend" is not marketing language that happens to sound technical. It is a specific allowance in federal labelling rules, and a brand that prints it is exercising a right rather than describing a recipe.

The allowance sits at 21 CFR 101.36(c). It exists because Congress, in the Dietary Supplement Health and Education Act of 1994, had to reconcile two things people wanted at once: full disclosure for buyers, and some protection for a formulator's work.

Macro crop, panel line reading Proprietary Blend with one milligram figure, matte paper, soft raking light
Figure 1 The two words that end the disclosure Everything indented beneath this heading has its individual amount withheld. The single figure at the right is the only quantity on offer. Disclosure permitted by 21 CFR 101.36(c). Illustrative crop.

What the rule requires, and what it lets go

Required

The blend must be named. Every dietary ingredient in it must be listed. That list must run in descending order of predominance by weight. And the blend's total weight must be declared.

Not required

The amount of any individual ingredient. That is the whole of it — one omission, and it happens to be the only number a buyer is trying to find.

Two-column diagram, cream ground, forest-green rules, left column ticked, right column struck through, clean type, no icons
Figure 2 Four disclosures required, one omission permitted Name, full ingredient list, descending order by weight and a combined total are all mandatory. The per-ingredient split is the single thing the rule leaves out. Illustrative summary of 21 CFR 101.36(c).

How to spot one in three seconds

Look for indentation. A blend is set as a heading with an amount, and its components sit indented underneath with no amounts of their own.

The names vary — "Focus Matrix", "Proprietary Complex", "Cognitive Blend". The structure does not: one figure at the right, several names beneath it, no second column.

The case for blends, made properly

The argument

A formulation is real work. Someone chose the ratios, tested them, and paid a contract manufacturer to make them reproducible. Publishing the exact split hands that work to any competitor with a purchase order.

Where it genuinely holds

It holds best for complicated botanical formulas with a dozen components, where the ratios are the intellectual property and there is no patent to protect them. Trade secrecy is the only shelter available, and a full panel dissolves it.

That is a serious argument and it deserves to be stated at full strength before it is answered.

Why it fails the buyer anyway

Descending order is a very weak constraint

Knowing that ingredient one outweighs ingredient two tells you almost nothing. In a 500 mg blend of six ingredients, the first could be 495 mg or it could be 90 mg, and both satisfy the rule identically.

The total is a ceiling, not a distribution

The one thing a blend total does establish is an upper bound: no ingredient in a 300 mg blend can exceed 300 mg. That is genuinely useful, and it is the basis of the check below.

Two stacked horizontal bars of identical length, cream ground, forest-green segments in very different proportions, no gridlines
Figure 3 Two products, one label Both bars total 500 mg across six ingredients in the same descending order, so both print an identical panel. The distributions are not comparable. Illustrative distributions, constructed to satisfy 21 CFR 101.36(c).

Two formulas that print the same label

Ingredient Formula A Formula B Printed on both panels
First listed 300 mg 120 mg listed first
Second 100 mg 110 mg listed second
Third 50 mg 100 mg listed third
Fourth 30 mg 90 mg listed fourth
Fifth 15 mg 50 mg listed fifth
Sixth 5 mg 30 mg listed sixth
Blend total 500 mg 500 mg 500 mg

Both columns are lawful. Both print the same panel. A shopper comparing the two has no way to tell them apart, and neither does a reviewer.

The check that actually works

Take the blend total. Take the first-listed ingredient — the heaviest one. Compare that total against the range research has used for that ingredient, in a population resembling the buyer.

Worked example

Research on alpha-GPC in healthy adults has run at roughly 250–630 mg a day: a 2024 trial in Nutrients at 315 mg using a Stroop task, a 2018 dose-comparison at 250 and 500 mg, and a 2023 trial in cyclists at 300 mg.

So a blend totalling 200 mg cannot contain 250 mg of anything. If alpha-GPC is listed first among five ingredients in that blend, its amount is under 200 mg and probably well under. The conclusion is arithmetic, not suspicion.

The number that does not belong in this check

Higher alpha-GPC figures circulate in marketing — trials at 800 and 1,200 mg a day. Those were run in people with diagnosed cognitive impairment: a different population, different baselines, different endpoints. Quoting them at a healthy adult borrows authority from research that was not about them. The range and its populations are set out in how much alpha-GPC the research actually used.

Simple number line, cream ground, forest-green shaded band and a separate marker outside it, minimal labels, no gridlines
Figure 4 A blend total below the research range settles the question Research on alpha-GPC in healthy adults has run at roughly 250–630 mg a day. A blend whose entire declared weight falls below that band cannot contain a single ingredient within it. Nutrients 2024 (315 mg, Stroop task); a 2018 dose-comparison (250 and 500 mg); a 2023 trial in cyclists (300 mg). Healthy adults.

What the check cannot conclude

A blend total comfortably above the research range proves nothing at all. It is consistent with a generous amount and equally consistent with a trace, padded out by five cheaper ingredients. The check is one-directional: it can rule a product out, never in.

What a trade secret actually protects

The formula-protection argument assumes a competitor cannot find out. In practice a contract analytical lab can quantify most actives in a capsule from a retail sample, and the manufacturers who make these products largely make each other's.

So the secrecy holds against shoppers and journalists more reliably than against competitors — which is the wrong way round if protection is the stated purpose.

Close crop, single capsule opened on a matte cream surface, powder spilled in a small arc, natural window light
Figure 5 What a retail sample gives up Most actives in a finished capsule can be quantified by a contract analytical laboratory from a purchased pack. Secrecy on a panel is not secrecy from the trade. Illustrative photograph. No product identified.

Where the argument survives, and where it does not

Situation Does the protection argument hold?
Twelve botanicals in an unusual ratio, no patent Reasonably — the ratio is the work
A blend built around one well-known active Weakly — the ratio is not the innovation
Three or four ingredients, all commodity No — there is nothing to protect
A blend where the first-listed ingredient is also the most expensive No — the incentive runs the other way

What a formula can do instead

Print every amount and compete on something a competitor cannot copy from a panel: sourcing, the specification behind each ingredient, testing, and the willingness to show the paperwork.

None of that is heroic. It is simply a different bet — that the buyers worth having are the ones who read the back of the pack.

Macro crop, fully itemised panel with a figure against every row, matte paper, soft window light, cream ground
Figure 6 The same panel with nothing pooled A fully itemised panel carries one amount per row. It takes the same space, costs nothing extra to print, and is checkable against published research ingredient by ingredient. Illustrative crop, formatted per 21 CFR 101.36.

Three questions worth emailing

What is the amount of each ingredient in the blend, in milligrams per serving?

If that is confidential, what is the amount of the first-listed ingredient alone?

Can you send the certificate of analysis for the lot number on this pack?

A brand that answers the first question has effectively abandoned the blend and could print it. One that answers the second has given up the only number that mattered. One that answers neither has told you something too.

Overhead shot, a phone flat on a worn kitchen table beside a supplement pouch, screen showing a short typed email, morning light
Figure 7 The question that costs nothing to ask Requests for a per-ingredient split and a lot certificate of analysis are ordinary customer-service enquiries. The reply, or its absence, is itself information. Illustrative photograph. No brand identified.

Standing in the shop

Three steps, and none of them require a phone.

  • Find the blend total and treat it as the ceiling for every ingredient inside.
  • Read the first name in the list. That is the heaviest component, and nothing else.
  • If the total is below the range research has used for that first ingredient, the pack has answered you.

The rest of the panel reads the same way it always does. That sequence is in how to read a Supplement Facts panel in about sixty seconds.

One place the same logic applies twice

Mushroom ingredients can be pooled inside a blend and vague on their own account, since "polysaccharides" is a broader category than the compound most people are looking for. Both problems are read off the same parenthetical, and that is covered in the one number to look for on a mushroom supplement.

Our own position

There is no blend line on Weal Focus: five ingredients, five amounts, each with its form printed beside it. That is a choice about disclosure rather than a claim about dose, and it can be verified in the same ten seconds as anyone else's panel.

Three-step flow diagram, cream ground, forest-green rules, three boxes left to right, generous white space, no icons
Figure 8 The whole check, in three steps Read the total as a ceiling, read the first name as the heaviest ingredient, compare the ceiling against the published range for that ingredient in a comparable population. Illustrative diagram. Ranges and populations on our citations page.

These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.

These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.

Sources for every figure, with the population studied in each, are on our citations page.