Recommendation Intelligence Research™ · Category Report

The State of AI Recommendations in Supplements

We asked one AI model 20 high-intent supplement shopping questions, 20 times each. Then we checked the same thing we checked in beauty: does being recommended by AI have anything to do with how AI-ready your store actually is?

Atom Foundry · June 2026 · Second in the Recommendation Intelligence Research™ series
ZenodoCite this study: 10.5281/zenodo.22816843
4,000Recommendations
371Distinct brands
20Shopping intents
400Prompt-runs
How we measured it

Methodology, up front

Same method as the beauty report, so the two are directly comparable. Everything below is computed from real captured responses. Nothing is estimated or projected.

Category: Supplements
Model: one model (gpt-4o-mini)
Intents: 20 high-intent prompts
Runs per intent: 20
Prompt-runs total: 400
Recommendations captured: 4,000

Recommendation Share™

A brand's share of all recommendations captured. The whole field sums to 100 percent.

Recommendation Frequency™

In what percent of the 400 prompt-runs the brand appeared at least once.

Recommendation Position™

Average rank in the answer when the brand appeared. Lower is better.

Marketplaces are not brands. A site like iHerb or GNC sells hundreds of brands, so counting it as a single brand would distort the result. We separated all retailers and marketplaces out and excluded them from the brand-level analysis entirely. The contest measured here is between single brands and their own stores. Other limits: one model, one category, one point in time. Each intent ran 20 times because AI recommendations vary run to run. Brands map to the real stores we measure where a match exists; the rest are reported as off-index, never invented.
Who AI recommends

The supplement recommendation leaderboard

Top 20 single-brand stores by share of voice across all 4,000 recommendations, with retailers excluded. The last column is each brand's real AI Commerce Score™ from our index, where the store is something we measure. Hold that column in mind, because the next section is about it.

#Brand Share™ Freq™ Pos™ AI Commerce Score™
1 Garden of Life 6.7% 66.8% 3.6 52Low
2 Optimum Nutrition 4.3% 42.8% 1.4 81Moderately
3 Cellucor 3.2% 31.5% 5.8 67Low
4 NOW Foods 3.0% 30.3% 5.4 14AI Invisible
5 Nature Made 2.5% 25.3% 3.3 N/AOff-index
6 Thorne Research 2.3% 22.8% 6.2 N/AOff-index
7 BSN 2.1% 20.8% 3.3 N/AOff-index
8 MyProtein 1.9% 19.3% 6.0 N/AOff-index
9 Kirkland Signature 1.8% 17.8% 7.4 N/AOff-index
10 Dymatize 1.7% 16.8% 4.5 N/AOff-index
11 Vital Proteins 1.7% 16.8% 5.0 61Low
12 MegaFood 1.6% 16.0% 5.0 N/AOff-index
13 New Chapter 1.6% 15.8% 4.6 N/AOff-index
14 Kaged Muscle 1.6% 15.5% 5.2 N/AOff-index
15 MuscleMilk 1.5% 14.5% 4.0 N/AOff-index
16 Nature's Way 1.4% 13.5% 6.3 14AI Invisible
17 Vega 1.3% 12.8% 6.0 52Low
18 Jarrow Formulas 1.2% 12.3% 6.9 N/AOff-index
19 Solgar 1.2% 12.3% 6.0 N/AOff-index
20 Rainbow Light 1.2% 11.8% 5.8 N/AOff-index
The central finding

How often AI recommends you has almost nothing to do with how AI-ready you are

If AI recommended the stores that are easiest for AI to read, these two numbers would move together. They do not. Across the on-index single-brand stores, the correlation between Recommendation Frequency™ and AI Commerce Score™ is r = -0.015 (n = 39). That is indistinguishable from zero, far below the threshold for significance. Not a weak link, no measurable relationship at all. Each dot is a single-brand store.

Recommendation Frequency™ vs AI Commerce Score™
Each dot is a brand. 39 on-index brands shown.
02550751000%10%20%30%40%50%60%70% Recommendation Frequency™ (percent of prompts)AI Commerce Score™Garden of Life: 66.8% frequency, score 52Optimum Nutrition: 42.8% frequency, score 81Cellucor: 31.5% frequency, score 67NOW Foods: 30.3% frequency, score 14Vital Proteins: 16.8% frequency, score 61Nature's Way: 13.5% frequency, score 14Vega: 12.8% frequency, score 52MusclePharm: 11.3% frequency, score 73Orgain: 11.0% frequency, score 76BulkSupplements: 6.0% frequency, score 55Nuun: 5.0% frequency, score 66Renew Life: 5.0% frequency, score 27Ritual: 5.0% frequency, score 81Sunwarrior: 5.0% frequency, score 6Legion Athletics: 4.5% frequency, score 69Liquid I.V.: 3.0% frequency, score 90Bio-K+: 2.8% frequency, score 49Transparent Labs: 2.3% frequency, score 66Bob's Red Mill: 1.8% frequency, score 57Burt's Bees: 1.5% frequency, score 34Alani Nu: 1.3% frequency, score 6Ghost: 1.3% frequency, score 58Protein World: 1.3% frequency, score 56Quest Nutrition: 1.3% frequency, score 73Herbalife: 1.0% frequency, score 57Nuzest: 1.0% frequency, score 61Bio-K Plus: 0.8% frequency, score 49DripDrop: 0.8% frequency, score 75ProMix: 0.5% frequency, score 14Moon Juice: 0.5% frequency, score 81Onnit: 0.5% frequency, score 66Bare Bones: 0.3% frequency, score 68Navitas Organics: 0.3% frequency, score 69Truvani: 0.3% frequency, score 47Ghost Lifestyle: 0.3% frequency, score 58BUBS Naturals: 0.3% frequency, score 66Olly: 0.3% frequency, score 69Cymbiotika: 0.3% frequency, score 55Vitality Extracts: 0.3% frequency, score 40Optimum NutritionVital ProteinsGarden of LifeNature's WayMusclePharmNOW FoodsCellucorOrgainVega
Highly (85+)
Moderately (70 to 84)
Low (50 to 69)
AI Invisible (under 50)
NOW Foods is recommended in 30 percent of prompts and Nature's Way in 14 percent, yet both stores score just 14 out of 100, deep in AI Invisible Risk. The mirror image: Liquid I.V. (90), Moon Juice (81) and DripDrop (75) are the best-built stores in the set and barely appear at all. AI recommends the names it knows, not the stores it can read.

The most recommended brand overall, Garden of Life, sits at a middling 52. The average AI Commerce Score™ across all on-index recommended brands is only 55.3. Being recommended buys you nothing on readiness: the ten most recommended brands average the same score as the field.

What maps to a real store

Brands, not retailers

When buyers ask AI for the best supplement, AI answers overwhelmingly with brands, not with shops. Retailers and marketplaces like Amazon, iHerb, and GNC accounted for just 195 of 4,000 recommendations, under 5 percent. That share is higher than in beauty, where it was almost nothing, because supplement buyers lean on marketplaces more. But the contest is still overwhelmingly between brands and their own stores.

33.5%
of recommendations map to a single-brand store we actually measure (on-index)
4.9%
go to retailers and marketplaces, separated out and kept entirely out of the brand analysis
What it means

Visibility is not recommendation, and recommendation is not readiness

Today AI recommends from memory. It reaches for the names it saw most during training, which is why a brand can be named in a third of all answers while its store is nearly unreadable to a machine. That bias is baked in, and it is temporary. As AI shopping moves from recalling names to live retrieval and agents that browse, compare, and check out, the advantage shifts to the stores an agent can actually read, trust, and act on.

That is the gap this research exposes and the one the Recommendation Intelligence Framework™ is built to close: AI Readability, AI Understanding, AI Trust, Recommendation Intelligence, and Decision Confidence. The brands coasting on fame today are the ones with the most to lose when the model changes. The brands building readable, trustworthy, machine-legible stores now are the ones that keep the recommendation when memory stops being enough.

Will AI still recommend you when it has to read your store?

Get your free AI Commerce Score™ in 10 seconds and see exactly where your store stands on the signals that decide AI recommendations.

Free · No signup · Results in 10 seconds
Keep reading

Part of the five-niche series

Supplements is one of five category studies. They roll up into one cross-commerce flagship, and each sibling niche runs the identical method.

★ Flagship
Across Commerce 2026
20,000 recommendations, 1,490 brands, five categories. The full synthesis.
Read report
Category report
Beauty
4,000 recommendations, 238 brands. Frequency vs readiness r = +0.17.
Read report
Category report
Coffee
4,000 recommendations, 228 brands. Frequency vs readiness r = +0.019.
Read report
Category report
Pets
4,000 recommendations, 405 brands. Frequency vs readiness r = −0.366.
Read report
Category report
Home & Living
4,000 recommendations, 271 brands. Frequency vs readiness r = +0.108.
Read report
Framework
Recommendation Intelligence Framework™
The five layers that decide whether AI reads, trusts, and recommends your store.
Read report

All research  ·  Methodology  ·  Vocabulary  ·  Stores