Share of voice across ChatGPT, Perplexity, Claude, and Google AIO — measured across 5.714 shopper prompts and 42 brands.
Different AI models weight different sources. A brand that wins ChatGPT may lag in Google AIO if its retail footprint is weak.
Ritual and Thorne together account for roughly 55.8% of all AI-generated brand mentions in wellness-supplement queries, with Ritual leading at 28.7% share of voice and an average citation position of 1.7. In a category tracking 42 D2C and legacy brands, two names absorb the majority of consumer-facing AI attention. For the remaining 40 brands, the aggregate ceiling is 44.2% — and 24 of them registered zero mentions across 240 engine calls.
The competitive dynamic is bifurcated at the top. Ritual and Thorne trade leadership depending on engine: Thorne leads ChatGPT (33.3% vs. Ritual's 28.3%), while Ritual leads Gemini (46.7% vs. 41.7%), Perplexity (21.7% vs. 16.7%), and Google AIO (18.3% vs. 16.7%). Athletic Greens (AG1) — at 13.8% SoV and tied with Moon Juice for the best average position (1.5) among brands with meaningful volume — sits a clear third by visibility. Below these three, the drop is severe: HUM Nutrition at 13.3% holds fourth, then a cliff to single digits.
A brand sitting between 10% and 30% SoV in this category should spend the next quarter doing three things: defending branded-comparison prompts, investing in high-authority health publisher and practitioner citations to compete on clinical and certification queries, and systematically testing where unclaimed high-intent prompt space exists. The economics of claiming empty prompt slots are materially better than displacing an incumbent — but the empty slots must be identified through prompt-level measurement, not assumed.
Wellness supplements is a category where purchase decisions are research-intensive and trust-constrained. Consumers ask questions before buying: is it third-party tested, does it contain sugar alcohols, is it better than the brand a friend recommended, is it worth the subscription cost. These are exactly the queries that AI assistants now answer directly, compressing what used to be a multi-tab research flow into a single generative response. When an AI engine names three brands in an answer to "best wellness supplement brands for overall wellness," it has already pre-filtered the consideration set before the shopper touches a product page.
In this sample, ChatGPT, Perplexity, Gemini, and Google AIO all surfaced brand citations, with Gemini producing the highest top-end mention rates (Ritual 46.7%, Thorne 41.7%) and Google AIO the narrowest spread. The implication is not that AI search has replaced Google or Amazon for supplement discovery — it has not — but that the citation graph forming now will determine which brands are defaults when volume does shift. Brands that wait until AI search is a measurable revenue channel will be competing to displace incumbents rather than to establish them.
| Brand | SoV % | Avg Position |
|---|---|---|
| Ritual | 28.7 | 1.7 |
| Thorne | 27.1 | 1.8 |
| Athletic Greens (AG1) | 13.8 | 1.5 |
| HUM Nutrition | 13.3 | 2.4 |
| Care/of | 8.8 | 2.3 |
| Moon Juice | 7.1 | 1.5 |
| Olly | 6.3 | 2.8 |
| Seed Health | 5.8 | 2.2 |
| Momentous | 5.4 | 2.1 |
| Huel | 3.3 | 2.4 |
| Gainful | 2.1 | 4.0 |
| Vital Proteins | 2.1 | 3.0 |
| Bloom Nutrition | 1.7 | 3.3 |
| Needed | 1.7 | 3.8 |
| Arrae | 1.3 | 2.0 |
| Four Sigmatic | 1.3 | 2.3 |
| The Nue Co | 0.8 | 6.0 |
| Perelel Health | 0.4 | 6.0 |
| Sakara Life | 0.4 | 8.0 |
| Cymbiotika | 0.4 | 3.0 |
| Recess Mood | 0.4 | 4.0 |
| Onnit | 0.4 | 5.0 |
| First Day | 0.4 | 1.0 |
| Lumen | 0 | — |
| Elix | 0 | — |
| Primal Queen | 0 | — |
| Dirty Lemon | 0 | — |
| MUD\WTR | 0 | — |
| Bulletproof | 0 | — |
| Ladder Sport | 0 | — |
| Goop | 0 | — |
| ZBiotics | 0 | — |
| Kin Euphorics | 0 | — |
| Modere | 0 | — |
| Swanson Health | 0 | — |
| Supergut | 0 | — |
| Magic Mind | 0 | — |
| Feals | 0 | — |
| Three Ships Beauty | 0 | — |
| Genexa | 0 | — |
| Hims & Hers | 0 | — |
| Function Health | 0 | — |
The gap between the #1 and #2 brands by SoV is narrow — 1.6 percentage points separates Ritual and Thorne. The gap between the #2 and #3 brands by SoV is 13.3 points. The top tier is effectively a duopoly, with AG1 and HUM competing for the clear-third slot at roughly half the mention rate.
Athletic Greens (AG1) punches above its SoV weight on position quality. Its average position of 1.5 ties for the strongest among higher-visibility brands (Moon Juice also sits at 1.5; First Day is lower at 1.0 but on negligible 0.4% SoV). When AG1 is mentioned, it is typically mentioned first — a pattern consistent with narrow category ownership. Moon Juice, at 7.1% SoV, combines lower overall visibility with a strong average position when it does appear. On the other end, Olly has 6.3% SoV but a 2.8 average position, placing it typically later in the recommendation set, around third on average, rather than as a primary recommendation.
Thorne leads at a 33.3% mention rate, followed by Ritual at 28.3%. HUM and AG1 follow at 16.7% and 15.0%. Thorne's edge on ChatGPT may reflect the weight ChatGPT places on practitioner-authored and clinical-certification content, though source-level attribution is not available from this dataset.
Ritual leads at 21.7%, Thorne at 16.7%. The narrower spread indicates Perplexity distributes citations across a wider brand set than the other engines in this sample.
Gemini produced the highest top-end mention rates of any engine: Ritual at 46.7% and Thorne at 41.7%, with AG1 at 16.7% and HUM at 21.7%. For top-tier brands, Gemini is the most concentrated citation environment in the sample — a meaningful channel, not a conservative one. For challengers, this concentration cuts the other way: displacement is harder where incumbents are cited nearly every time.
Ritual leads at 18.3%, Thorne at 16.7%, AG1 at 11.7%. Notably, Moon Juice ties HUM at 6.7% — a stronger showing on this engine than its overall category SoV would predict.
Prompt-level outputs were not part of the structured dataset underpinning this report, so the clusters below are framed as category hypotheses that a follow-on measurement pass should validate. Based on the engine-level patterns observed:
The operational point: a brand in this category should commission prompt-level measurement against these five clusters before committing budget, not after.
A prompt-level null analysis was not available in the source dataset for this report. That said, the category logic of unclaimed territory still applies and should frame the next measurement cycle. Moving into an empty prompt slot is materially cheaper than displacing Ritual or Thorne on a contested prompt, because the incumbent's citation graph does not yet exist. The five prompt archetypes most worth testing for null or weak results in a follow-on run:
These should be treated as hypotheses for the next measurement cycle, not as confirmed empty slots.
For a brand currently in the 10-30% SoV band — Ritual, Thorne, AG1, and HUM sit in this range — the next 90 days should focus on three concrete moves.
Move 1: Earn placements in high-authority health publishers within 30 days. In a YMYL category, retrieval systems weight established medical and consumer-health domains (Healthline, Verywell, Examine, Medical News Today, WebMD) far above Substack or Medium. The tactical program is a digital-PR push targeting credentialed-author review inclusions on those domains, paired with on-site schema-marked clinical validation (ingredient studies, third-party testing certificates, credentialed medical reviewers). Practitioner-authored Substacks can supplement this program but should not be the center of it.
Move 2: Build a branded-comparison content asset within 60 days. Incumbents win "alternatives to [incumbent]" queries in part because they have seeded comparison content themselves. A challenger should publish a head-to-head comparison hub covering its top five named competitors, with honest trade-off analysis and structured data. The goal is not to win every comparison — it is to be named in every comparison.
Move 3: Claim one unclaimed prompt cluster within 90 days. Commission a prompt-level scan against the five archetypes above, pick the cluster with the weakest incumbent citation graph, and over-invest. For a use-case cluster like "energy and focus," this means one clinical white paper, three dietitian-authored articles placed on tier-1 health domains, two Reddit AMAs with credentialed staff, and a dedicated landing page with schema markup. The target is not Google organic ranking — it is citation density across the sources that ChatGPT, Perplexity, and Gemini retrieve from. A brand willing to spend against a single empty cluster for one quarter can plausibly own it before a larger competitor notices.
This analysis is based on 240 engine calls distributed across ChatGPT, Perplexity, Gemini, and Google AIO, implying approximately 60 shopper prompts across four engines, designed to mirror real D2C query distributions in the wellness-supplement category (category discovery, branded comparison, ingredient and certification queries, use-case queries, and retailer-channel queries). Share of voice is computed as the percentage of prompts in which a brand was named in the AI response. Average position is computed across all mentions for a given brand. 42 brands were tracked in total. Results reflect a single measurement window and should be re-run quarterly to capture drift in engine citation behavior. Prompt-cluster and unclaimed-territory conclusions in this report are framed as hypotheses pending a follow-on prompt-level measurement pass.
Seeno tracks how your brand gets named by ChatGPT, Perplexity, Claude, and Google AIO — automatically, across your category's real shopper prompts.
Run a free audit →Free · no signup · 3-minute report