Share of Voice in AI Search: How D2C Brands Are Losing Ground
We ran 4,800 AI queries across ChatGPT, Gemini, and Perplexity. Most brands have 0% share of voice.
The research
Between January and February 2026, Vialtry ran 4,800 product recommendation queries across ChatGPT (GPT-4o), Gemini 1.5 Pro, and Perplexity across 8 D2C categories. Each query was designed to match real buyer intent — the way actual shoppers phrase product requests to AI assistants.
The headline finding
Median AI Share of Voice across all brands and categories: 0%.
The majority of D2C brands — including many with strong Google rankings, significant ad spend, and healthy revenue — appeared in zero AI-generated product recommendations across 600 queries in their category.
The correlation between Google search visibility and AI recommendation visibility is weak. These are different games with different rules.
Category breakdown
Skincare showed the most differentiated results. Approximately 8% of brands captured the majority of AI recommendations. These brands shared one characteristic: highly structured, attribute-complete product data with clear certifications and use-case targeting.
Supplements showed similar winner-take-most dynamics. Brands with NSF or Informed Sport certification prominently structured in their data appeared in a disproportionate share of supplement queries. Brands without third-party testing visibility were largely absent.
Apparel was the most challenging category. Even well-known apparel brands showed low AI Share of Voice due to weak attribute data.
Home decor showed patterns around dimension data. Brands with precise, structured dimension information appeared significantly more in use-case-specific queries.
Platform differences
ChatGPT was most brand-conservative — recommending fewer specific brands, weighting structured data quality most heavily.
Gemini showed stronger correlation with Google Shopping data quality. Brands with well-optimized Merchant Center feeds had meaningfully higher Gemini SOV.
Perplexity was most accessible for mid-tier brands. On-page schema and editorial mentions have more direct influence. Several smaller brands with strong structured data appeared regularly despite minimal ChatGPT or Gemini visibility.
What the top-quartile brands did differently
Across all categories, top-quartile brands shared four characteristics:
1. Complete, current schema.org Product markup with accurate price and availability 2. Structured category-specific attribute data in discrete, parseable fields 3. Third-party certifications properly marked up as attributes 4. Off-page presence — editorial coverage, comparison articles, review site mentions
None of these are difficult to achieve. They require systematic catalog work, not design changes or ad spend.
See your brand's current AI Share of Voice across ChatGPT, Gemini, and Perplexity — free at vialtry.com/grader.