Share of voice across ChatGPT, Perplexity, Claude, and Google AIO — measured across 34.667 shopper prompts and 15 brands.
Different AI models weight different sources. A brand that wins ChatGPT may lag in Google AIO if its retail footprint is weak.
Olipop is named in 82.3% of the 130 shopper prompts we tested across ChatGPT, Perplexity, Claude, and Google AIO, and it holds an average mention position of 1.2. It shows high visibility across every major AI engine, with engine-level rates ranging from 77.4% on Perplexity to 85.6% on ChatGPT. For a consumer asking an LLM about healthy soda, Olipop is the leading default recommendation.
The competitive structure is top-heavy. Olipop (82.3%), Poppi (72.3%), and Culture Pop (43.5%) are the three brands most frequently named across prompts. (Note: SoV here is prompt incidence, not an additive market share; multiple brands are cited per answer.) Below them, Health-Ade (15%) and De La Calle (10.4%) are cited occasionally; five of the fifteen tracked brands returned zero mentions, and another four sit below 3.1% SoV, with Spindrift at exactly 3.1%. One plausible mechanism behind the top two's lead is a feedback loop between press coverage, retail ubiquity, and AI citation, but this dataset does not test that hypothesis directly.
A brand in this category with low-double-digit share of voice should not attempt to displace Olipop or Poppi on head terms in the next quarter. The more defensible path is directional: invest against functional and comparative query territory (post-workout, inulin vs. chicory, prebiotic soda vs. kombucha) where the top three are less entrenched. These prompts should be cheaper to win and they map to higher-intent purchase moments, though the timeline for citation assets to compound into head-term visibility is uncertain and depends on engine-specific retrieval and retraining dynamics.
Prebiotic soda is a category built on a question: does this actually do something for my gut? That question is well-suited to generative search and poorly suited to traditional SERPs. A shopper deciding between a can of Olipop and a can of Poppi wants a paragraph, not ten blue links. They are asking LLMs whether the fiber dose is meaningful, whether the sweeteners are tolerated, and whether the taste is acceptable relative to legacy soda. Every one of those questions resolves to a brand recommendation inside the answer.
The shift is early. AI-driven discovery is still a minority of category search volume, and most purchases still originate at retail shelf or from Instagram. But the behavior is directional: prebiotic soda skews to a younger, health-literate buyer who treats ChatGPT and Perplexity as a second opinion before trying a new functional beverage. The brands that are cited today are shaping the consideration set for the cohort that will define the category over the next 24 months. This report measures who is winning that citation layer right now.
| Brand | SoV % | Avg Position |
|---|---|---|
| Olipop | 82.3 | 1.2 |
| Poppi | 72.3 | 1.6 |
| Culture Pop | 43.5 | 2.5 |
| Health-Ade | 15.0 | 4.1 |
| De La Calle | 10.4 | 4.3 |
| Spindrift | 3.1 | 4.0 |
| Bubly | 1.5 | 1.5 |
| Remedy Drinks | 1.2 | 6.0 |
| United Sodas of America | 0.8 | 6.0 |
| Sanzo | 0.4 | 7.0 |
| Koe Kombucha | 0 | n/a |
| Lemon Perfect | 0 | n/a |
| Ugly Drinks | 0 | n/a |
| Flying Embers | 0 | n/a |
| Moment | 0 | n/a |
The tracked set deliberately includes adjacent beverages (sparkling waters, kombuchas, tepache, functional waters) to test how porous the category is in LLM answers. Conclusions about prebiotic-soda competition below should be read against the core category participants (Olipop, Poppi, Culture Pop, De La Calle, Health-Ade's SunSip line, United Sodas, Remedy); the rest are adjacency probes.
The gap between Olipop and Poppi is 10.0 percentage points of SoV but only 0.4 positions on average rank. That is consistent with the two brands being the models' default pair, with Olipop listed first slightly more often. The gap between Poppi and Culture Pop is 28.8 points, and the gap between Culture Pop and Health-Ade is another 28.5 points. Three brands define the category; everyone else is an asterisk.
One plausible explanation for Olipop's lead is heavier editorial and retail coverage in sources LLMs tend to ingest for CPG recommendations, but this dataset does not attribute mentions to specific source types, so this remains speculation rather than finding. Poppi's broad consumer visibility may be reinforced by major brand moments such as its Super Bowl spend; we cannot confirm that mechanism from this data. Culture Pop materially over-indexes in this dataset relative to the rest of the non-top-two field, which is notable given its smaller marketing footprint, though without external benchmarks on spend and distribution we cannot make a definitive "punching above its weight" claim.
Bubly deserves one line of its own. At 1.5% SoV with an average position of 1.5, it is likely being pulled into answers when a prompt touches mainstream or zero-sugar sparkling water, and when it appears, it ranks high. That is not prebiotic-soda visibility. It is a category-adjacency artifact, and it is one reason the methodology intentionally tracked adjacent brands.
The engine-level rates are real; the underlying reasons are hypotheses worth testing rather than conclusions this dataset can prove.
Olipop (85.6%) and Poppi (83.7%) are effectively tied, with Culture Pop a strong third at 65.4%. ChatGPT shows the broadest co-citation pattern across the top three in the category.
Olipop leads at 77.4%, Poppi at 64.5%, Culture Pop drops to 25.8%. The Culture Pop fall-off may reflect Perplexity's citation-first architecture weighting fresher or more linkable sources, though that is a hypothesis, not a finding.
Claude is the least concentrated among the top three: Olipop 81.0%, Poppi 76.2%, Culture Pop 71.4%. Spindrift (14.3%) also surfaces here more than elsewhere, suggesting broader better-for-you beverage coverage may be in play.
Olipop 80.8%, Poppi 62.5%, Culture Pop collapses to 21.2%. Google AIO is the most concentrated engine in the category. This is where the long tail of the tracked set has the least chance of being cited, though attributing the concentration to any specific retrieval mechanism would overstate the evidence.
The test set was constructed across five prompt clusters: category-definition, functional-benefit, ingredient-integrity, taste/experience, and channel/availability. Cluster-level performance is not reported in the aggregate dataset above, so the observations below are drawn from external category knowledge and should be read as hypotheses to validate with prompt-level logs, not as findings supported by the SoV numbers.
Category-definition queries ("best prebiotic soda brands 2025", "top healthy sodas with probiotics"). The head-term battleground. Given aggregate SoV, this is almost certainly the cluster most dominated by Olipop, Poppi, and Culture Pop.
Functional-benefit queries ("best prebiotic drink for bloating", "which prebiotic soda actually works"). The cluster with the highest purchase intent, and the one where dietitian-style editorial coverage matters most.
Ingredient-integrity queries ("prebiotic soda without fake sweeteners", "best low-sugar prebiotic soda"). Clean-label positioning is most relevant here; Culture Pop's market narrative aligns with this cluster but we have not measured it.
Taste and experience queries ("prebiotic soda that tastes good"). Poppi's flavor-first brand narrative in consumer media is structurally aligned with this cluster.
Channel and availability queries ("top-rated prebiotic drinks on Amazon", "prebiotic soda brands at Whole Foods"). Retail footprint should favor Olipop and Poppi; Culture Pop's narrower retail distribution likely produces uneven reads across sub-queries.
A follow-on analysis with prompt-level logs would confirm or disconfirm these reads. They are the right next cut of the data, not a conclusion of this one.
The aggregate dataset does not include prompt-level outputs, so we cannot name specific zero-mention prompts from this study. What the data does support is a structural observation: outside the top three brands, citation frequency collapses sharply, and specific query types in generative search tend to produce low branded output for reasons that matter strategically.
Informational and ingredient-education queries (for example, "inulin vs chicory root") often resolve without a brand mention because the LLM can answer the biochemistry without recommending a product. That is not the same as a content gap a brand can simply fill. However, it does mean that high-quality, scientifically grounded content on these topics measurably increases the probability that an LLM will surface a brand as a concrete example alongside the biology explanation.
Three prompt archetypes where the combination of commercial intent and sparse top-three dominance makes the economics most favorable for a challenger brand:
A follow-on study with prompt-level logs would quantify which of these are genuinely zero-mention today. The directional read is that challenger economics are better in these archetypes than on head terms.
For a brand in this category currently holding roughly 10-30% SoV, three directional moves over the next 90 days. The obvious readers are De La Calle (tepache, adjacent to but not native to prebiotic soda) and Health-Ade, which should apply this playbook specifically to its SunSip sub-brand rather than its legacy kombucha line; driving prebiotic-soda queries to kombucha URLs will confuse both consumers and LLM retrieval.
Move 1: Fund credentialed, earned-media content on ingredient education, not sponsored placements. Commission a registered dietitian to author a rigorous explainer on inulin vs. chicory root vs. cassava fiber, and pursue two parallel distribution paths: (a) pitch the RD and a data-backed angle to independent writers at Healthline, Verywell, or EatingWell as earned media, and (b) pay the RD to publish on their own high-authority domain (a popular dietitian Substack, for example). Tier-one editorial sites cannot be bought for $10K; sponsored content on those domains is tagged as such and is typically devalued by LLM retrieval. The earned-media plus credentialed-creator path is the realistic version of this play. Budget: $10-20K for the RD, study design, and pitching support.
Move 2: Seed a structured comparison asset for "prebiotic soda vs. kombucha" on Reddit and in two long-form Substack newsletters covering functional beverages. The working assumption is that Reddit and independent newsletter content are part of what LLMs retrieve and ingest in CPG answers; we do not have source-level citation data to confirm weighting. The goal is not a viral post; it is a durable, high-quality thread and newsletter series that accumulate links and signals over time. Treat timelines to LLM visibility as uncertain and dependent on each engine's retrieval and retraining cadence.
Move 3: Build a post-workout recovery landing page anchored by a consumer perception study of 50-100 users (hydration and GI-comfort self-report) and pitch it to three outlets: Well+Good, mindbodygreen, and Prevention. Be honest about what this is: a scrappy, proprietary consumer perception study that gives wellness-vertical journalists a concrete data hook. It is not a clinical trial and should not be positioned as one. (Seed Probiotics' editorial footprint rests on peer-reviewed human clinicals, a scientific advisory board, and the SHIME model; that is a different, far more expensive playbook and not the one being proposed here.) The hypothesis is that proprietary data increases publisher pickup, which indirectly improves LLM visibility over subsequent retraining cycles.
Across all three moves, the common principle: do not fight for the head term. Invest against prompt territory where the top three are less entrenched, produce genuinely credible citation assets, and accept that the compounding effect on the leaderboard is probabilistic, not guaranteed.
This analysis is based on 520 engine calls across ChatGPT, Perplexity, Claude, and Google AIO, executed against 130 shopper prompts designed to mirror the real query distribution a prebiotic-soda buyer produces across the consideration funnel (category-definition, functional-benefit, ingredient-integrity, taste, channel, and comparison queries). Share of voice is computed as the percentage of prompts in which a given brand was named in the generated answer; it is not an additive market share and brand SoVs do not sum to 100. Average position is the mean ordinal position of the brand's mention across the prompts where it appeared. Fifteen brands were tracked, intentionally including adjacent beverages (sparkling water, kombucha, tepache, functional water) to measure the porosity of the category in LLM answers; five returned zero mentions and are reported as such rather than excluded. Prompt-level outputs, cluster-level attribution, and source-level citation breakdowns are not part of this dataset; any claims drawing on those inputs are flagged as hypotheses rather than findings.
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