Survey: AI Search Visibility of 30 Japanese EC Brands — Mid-size and Small Brands Named in 47% of Industry Questions vs. 34% for Large Brands
Big-O Information Intelligence Corp. asked four generative AI systems 1,200 questions about 30 Japanese EC / D2C brands. In industry-level questions that did not name any brand, mid-size and small brands were named more often (47%) than large brands (34%).
— Big-O Information Intelligence Corp.
On 2026-10-08, Big-O Information Intelligence Corp. released via ValuePress the results of a survey covering 30 Japanese EC / D2C brands (10 large, 10 mid-size, 10 small). A total of 1,200 questions were put to ChatGPT, Gemini, Claude and Perplexity. Across the 480 industry-level questions that did not mention any brand name, mid-size and small brands were named in 47% of answers, ahead of large brands at 34%.
Background
Consumers and business partners increasingly ask generative AI directly for "recommended brands" or "environmentally responsible companies". Even when a company publishes its environmental policy and CO₂ reduction measures on its website, there is no guarantee that generative AI understands and cites that information correctly. The survey was conducted to understand how the information companies publish is recognised and cited by generative AI.
Key findings
1. In industry-level questions, small and mid-size D2C brands are named more often than large brands: across 480 questions asked by industry without naming a brand, mid-size and small brands were named in 47% of answers, versus 34% for large brands.
2. AI can "talk about" sustainability: 88% of questions received a concrete answer. Of 600 questions that named a brand and asked about its sustainability initiatives, CO₂ reduction targets, carbon-neutral / GX policy or environmental certifications, AI gave concrete content in 526 (88%).
3. But 17% of those did not cite the official website: of the 526 concrete answers, 90 (17%) did not cite the company's official website as a source.
4. Six companies had no sustainability information findable on their official site, yet AI described their initiatives: for 6 of the 30 companies, the survey method found no sustainability / CO₂ / GX statements on the official website, yet AI still described the company's initiatives. All six are small D2C brands.
5. GX recognition differs by company size: large 79, mid-size 44, small 28 (out of 100). Official-site disclosure volume showed the same gap (on-site sustainability information score: large 44, mid-size 12, small 6, out of 65).
6. Recognition rate by item: concrete CO₂ reduction target 52%; carbon-neutral / GX policy 58%; environmental certifications (ISO 14001, SBT, CDP, etc.) 52%.
7. Comparison across the four AI systems (1,200 answers): Gemini citation URLs (Google redirect format) were resolved to the destination origin domain before aggregation.
8. Three groups by AI search visibility × GX recognition: sustainable brands trusted by AI, 11; visible to AI but weak sustainability information, 16; not visible to AI, 3.
The "information gap" that follows SEO
The survey confirmed both cases: published information that is not cited, and unpublished information that AI "fills in" and narrates. Big-O's view is that companies need not only to publish information, but to put it in a state where AI can discover it, understand it and cite it as a trusted source.
Survey overview
Survey name: Japan EC Brand AI Search × GX Visibility Survey 2026 (1st edition)
Subjects: 30 Japanese EC / D2C brands (10 large, 10 mid-size, 10 small)
Period: 5–6 September 2026
AI systems: ChatGPT (with search), Gemini (search grounding), Claude (web search), Perplexity
Questions: 10 per company (5 direct brand questions, 4 industry questions, 1 competitor question), 1,200 answers in total
Analysis items: brand mention, recommendation, official-site citation, third-party citation, sustainability recognition, CO₂ recognition, GX recognition, certification recognition, consistency across AI systems
Method: the GX Recognition module of Big-O "EcoGEO Audit" (automatic answer classification + official-site disclosure scan)
Note: the survey does not assess any company's actual environmental performance or ESG rating; it analyses how published information is recognised and cited by generative AI. AI answers vary with question wording and timing. Company names are not disclosed individually.
Next: a 30-day public verification
From the release date, Big-O starts a "30-day public verification" to observe whether generative AI's recognition of its own brand changes after this release is distributed. The baseline (Day 0) is the 2026-10-05 (Monday) reading from the company's weekly Monday measurement. Progress and results will be published, including the case where nothing changes.
Check your own AI search visibility for free
Using the same perspective as this survey, you can check for free how generative AI sees your website (LLMO diagnosis, no login required). Results are produced by automated analysis and do not guarantee mention or recommendation by AI.
Free LLMO diagnosis: https://bigomorfi.tech/ja
About Big-O
Company: Big-O Information Intelligence Corp.
Location: Neihu District, Taipei, Taiwan
Founded: November 2023
Representative: Elaine Chang
Business: development and provision of the Big-O AI Brand Infrastructure Platform