What Each Nutra Niche Leans On: A Composition Map of 56,017 Beats

Do nutra niches actually build their copy differently?

Yes, and the gap is large enough to plan around. We pulled 56,017 extractions from transcribed VSLs and advertorials across 21 supplement niches, then indexed each niche's share of specific copy beats — hook, avatar, pain, promise, social proof, CTA, urgency — against the corpus-wide share for that same beat. An index of 1.00 means a niche uses that beat at exactly the average rate. A 1.25 means it leans on that beat 25% harder than the rest of the corpus combined. Read the index, not the count.

Fourteen niches cleared the 500-extraction floor we set for characterization; the other seven, folded into the 21-niche total, didn't generate enough transcript volume to trust an index score, so we left them out rather than report a number built on a handful of transcripts. Weight-loss alone accounts for 15,729 of the 56,017 extractions — more than a quarter of the entire corpus. Nerve (6,473) and memory (6,458) are a distant second and third. Diabetes sits mid-table at 3,408.

None of that ranks market size. It ranks what we transcribed most, which is a function of the ad spend we could observe and how many angles a category runs through before an offer goes dark. Say it plainly: weight-loss leads the extraction table because we sourced most heavily there, not because it is definitionally the biggest market in nutra. That's sourcing, not market size. The index scores are the more honest measure of what's actually different niche to niche.

Which niche leans hardest on social proof?

Weight-loss, by a clear margin among the niches we could characterize. Its social proof beat runs at 1.19 against corpus average, its avatar beat at 1.25, promise at 1.11, pain at 1.09. That's four beats over-indexed at once, which is unusual — most niches in our data pull hard on one or two beats and sit near average on the rest. That's rare in this data.

Read the avatar number first. An avatar beat is the section of copy where the reader is supposed to recognize themselves — age, body type, daily frustration, the 'this is you' moment before the pitch starts. At 1.25, weight-loss spends more real estate building that mirror than any other niche in the set. Social proof follows the same logic: testimonials, before/afters, reviewer counts, the visible weight of other buyers' experience. A niche doesn't over-index on both unless sellers believe the reader needs convincing on two fronts — that the product is for someone like them, and that it worked for someone like them.

That's what saturation looks like on paper. When dozens of offers make the same promise, the burden of proof shifts from does this work to does this work for someone who looks like me, and can you show me the receipts. If you're asking which supplement niche is most competitive, that shift — not the extraction count — is the honest place to look for an answer, and on that measure weight-loss is it. Weight-loss carries the largest promise and pain indices too, but neither is the standout; the standout is how much weight the niche puts on identity and evidence rather than the claim itself.

Why does diabetes over-index on hooks and CTAs?

Diabetes copy spends its effort at the two ends of the funnel, not the middle. Hook sits at 1.38 against corpus average — the single largest over-index of any beat we measured, in any characterized niche. CTA follows at 1.22, avatar at 1.19, urgency at 1.09. Social proof and promise, by implication, sit closer to average; the corpus doesn't give us their exact index here, so we won't guess at a decimal we don't have.

A niche that front-loads the hook and back-loads the CTA this hard is a niche where the copy is built to catch attention fast and close before the reader has time to think it through. Hook carries the weight here. Diabetes is one of the more heavily scrutinized health categories on both Meta and Google, and it sits inside the FTC's enforcement priorities as a matter of public record — the FTC's Health Products Compliance Guidance names blood-sugar and diabetes-management claims explicitly as an area drawing outsized attention. Under that kind of pressure, the safest place to put persuasive weight is often the opening line and the close, not a sustained proof section that has to survive a compliance review months into the offer's life.

It's a reasonable read of the pattern, not a confirmed cause. We can't confirm the mechanism. We can say the skew exists and describe what it looks like on the page; we can't say from index scores alone that regulatory pressure is driving it, which would need transcript-level analysis this dataset doesn't give us.

What does a proof-heavy niche imply about the buyer?

It implies a buyer who has been burned before, or at least who has seen enough competing claims to be skeptical by default. Weight-loss's 1.25 avatar and 1.19 social proof indices point at a reader who needs to see herself in the ad and see evidence before she'll act, not a reader who moves on a hook alone.

Contrast that with what the diabetes numbers imply about its buyer: someone who responds to a strong opening and a clear next step, with less copy apparently spent making the case in between. That's not a claim about which buyer is smarter or more cautious. It's a claim about what the seller believes will move each of them, distilled from tens of thousands of ads that had to survive real spend to get transcribed in the first place. Trust gets earned twice in weight-loss. Ad copy that doesn't convert gets pulled fast; what's left in a corpus this size is copy that worked well enough to keep running.

The practical read: in weight-loss, an offer with a thin testimonial section is fighting the market's default expectation, whatever the product itself claims to do. In diabetes, an offer that opens soft and builds slowly is doing the opposite of what the rest of the category has converged on. Neither pattern proves the underlying product claims are true. It proves what the copy has to do to get someone to click.

Where does the corpus have too little data to say?

On twelve of the fourteen characterized niches, and on seven niches entirely. We have index scores — hook, avatar, pain, promise, social proof, CTA, urgency — for exactly two niches in this dataset: weight-loss and diabetes. Nerve (6,473 extractions), memory (6,458), joint-pain (3,676), erectile-dysfunction (3,333), prostate (2,723), hearing (2,719), skin (1,011), lymphatic (863), vision (733), dental (618), gut (598), and lung (575) all cleared the 500-extraction floor for volume reporting, but we don't have their beat-composition indices to publish here. Volume isn't composition. Knowing nerve generated 6,473 extractions tells you how much we transcribed; it tells you nothing about whether nerve copy leans on urgency or proof.

Below the 500-extraction floor sit seven more niches folded into the 21-niche total, unnamed here because we don't trust a percentage built on a small sample. That's a deliberate omission, not an oversight — reporting an index off a thin stack of transcripts would produce a number that moves with the next batch we add. We'd rather say nothing.

The honest position is that this composition map currently covers two niches well and twelve niches by volume only. Treat any claim about nerve, memory, or ED copy leaning one way or another as a gap to fill, not a finding to act on yet.

How would you use a composition map before entering a niche?

Match your asset list to the skew before you brief a creative team, not after the first batch underperforms. If you're entering a niche that over-indexes on avatar and social proof the way weight-loss does, budget for multiple avatar variants and a real testimonial pipeline before launch — a single generic hook with no proof section is fighting the category's default, and it will surface as weak conversion in your own testing before you know why. Brief accordingly.

If the niche looks more like diabetes, with hook and CTA carrying the weight, put your production time into opening lines and closes, and don't over-invest in a long mid-funnel proof section the market apparently doesn't reward as heavily. Test it yourself. Index scores describe what's been running; they don't guarantee what converts for your specific offer.

• Pull the niche's index profile before writing a brief, if you have one — treat it as a starting allocation of creative effort, not a final answer.
• Where you don't have an index (twelve of our fourteen characterized niches, today), default to testing avatar, proof, and hook variants roughly evenly rather than assuming an undocumented skew.
• Don't confuse a niche's extraction volume with its competitiveness. A niche with less data here ran fewer angles in our sourcing, not necessarily fewer dollars in the market.

None of this replaces watching what's actually live right now. A composition index shows the historical shape of a niche's copy; it doesn't tell you which specific offer inside that niche is fresh versus already worn out, and that distinction still has to come from current tracking, not from an average built across a full corpus.

Frequently asked questions

Which supplement niche is most competitive, based on this data?

Weight-loss, based on composition rather than raw volume. Its avatar and social-proof beats run 25% and 19% above corpus average, the largest combined over-index of any niche we could characterize — a sign sellers there work harder to build identity and evidence than sellers in any other measured category.

Does higher extraction volume mean a bigger market?

No — extraction volume in this corpus reflects what we transcribed, not market size. Weight-loss leads the table at 15,729 extractions because we sourced most heavily there, out of 56,017 total across 21 niches. Treat volume as a sourcing artifact, not a demand signal, and look at the index scores instead.

What does a composition index of 1.00 mean?

It means a niche uses that copy beat at exactly the corpus-wide average rate, no more and no less. An index above 1.00 shows over-reliance relative to the rest of the corpus; 1.38, diabetes's hook index, means hooks appear 38% more than the average niche uses them.

Why does diabetes copy lean so hard on hooks and CTAs?

The pattern likely reflects the heavier compliance scrutiny diabetes claims face, per the FTC's Health Products Compliance Guidance, though the index data alone can't confirm the mechanism. Front-loading the hook and closing hard on the CTA needs less sustained mid-funnel proof to survive review than a long testimonial-driven middle section does.

How many niches have published composition index scores?

Two: weight-loss and diabetes. Twelve more niches cleared the 500-extraction reporting floor and have known volumes, from nerve at 6,473 down to lung at 575, but no published index breakdown here. Seven additional niches fell below the threshold entirely and aren't characterized in this release.

Sources

Named rather than linked — verify before relying on any figure below.

• FTC Health Products Compliance Guidance
• FTC Endorsement Guides
• Meta advertising policies on personal health claims