AI SEO / AEO Audit — Two Brothers Organic Farms (USA Store: twobrothersfood.com)

0. Executive Summary

Two Brothers Organic Farms' USA store is unusually well-positioned for AI-driven commerce already: it sits on Shopify's agentic-commerce stack and auto-exposes llms.txt, agents.md, a Universal Commerce Protocol (/.well-known/ucp) endpoint, and product pages with strong Product + FAQPage + AggregateRating schema (4.43/251 reviews on the hero ghee). AI crawlers are fully allowed in robots.txt, and Core Web Vitals are "good" on mobile.

The gap is not infrastructure — it's content and entity clarity:

1. Brand-name inconsistency on the US store. The homepage and key category copy repeatedly call the brand "Two Brothers India Farms" (14 occurrences on the homepage) while the legal/about/contact pages say "Two Brothers Organic Farms" and the domain/footer say "USA Two Brothers Farms." AI engines resolve entities from consistent naming; this fragmentation dilutes brand citation and confuses "who is this."
2. No H1 on the homepage — the single most important definitional element for AI answer-extraction is missing.
3. A major top-of-funnel visibility gap: TBOF owns its branded "best A2 ghee two brothers" queries (positions 1/2/8) but is completely absent from the generic, high-volume definitional query "what is A2 bilona ghee," which is dominated by competitors (Sri Sri Tattva, Amritangan, Gavyamart, Milkaaru, etc.). That is exactly the query AI Overviews and ChatGPT fan out to for ghee buyers.
4. Third-party comparison pages dominate "vs" queries (Authentic Urban, Heroveda, Shahji, Raghas) and TBOF has no "Two Brothers vs [competitor]" content on the US store to get cited from. Competitors are shaping the narrative.
5. Pricing is opaque to agents/citation: /pricing.md 404s, and the /policies/ path (shipping/refund) is Disallowed in robots.txt, so AI systems literally cannot read TBOF's shipping/return promises even though those pages exist (HTTP 200).
6. Thin, fragmented FAQ and no comparison/guide content in the US content set (only 2 "why ghee" pages exist; 83 total pages are mostly collections/product stubs).

Bottom line: The plumbing is ahead of most D2C brands. Closing it requires (a) entity-consistency fixes, (b) a definitional + comparison content layer, and (c) three small machine-readable file additions. Estimated impact: materially higher AI citation share on the highest-intent ghee/organic-pantry queries within 1–2 quarters.

1. Methodology & What Was Verified

Schema was verified by fetching raw HTML and parsing <script type="application/ld+json"> blocks — not via cleaned/markdown extraction, which would have stripped the scripts and falsely reported "no schema."

2. Technical Foundation — Strong (The Good News)

This is genuinely ahead-of-curve for a mid-size D2C food brand:

• AI crawler access is fully open. GPTBot, ChatGPT-User, PerplexityBot, ClaudeBot, anthropic-ai, Google-Extended, and Bingbot are all allowed. No accidental blocking.
• Agentic-commerce endpoints are live. The store exposes a valid llms.txt/agents.md describing UCP capabilities, a working /.well-known/ucp merchant profile (UCP 2026-04-08), and an MCP endpoint (/api/ucp/mcp). This means an AI agent can already discover, price-check, and transact — TBOF is ready for the "buy-for-me" agent wave.
• Product schema is excellent. The hero ghee (amorearth-desi-cow-a2-ghee) carries Product + Offer ($35 USD, InStock) + AggregateRating (4.43 across 251 reviews) + FAQPage + Brand. This is exactly the structured, citable data non-Google AI engines reward.
• Blog posts carry BlogPosting schema with author/date structure.
• Performance is healthy (CrUX: phone LCP ~970ms, INP ~180ms, CLS ~0.01).
• sitemap.xml is complete and includes a dedicated sitemap_agentic_discovery.xml plus blog/product/collection/page sitemaps.

No action needed on the agentic/plumbing layer — keep it, maintain it, and extend it.

3. AI Visibility Audit (Current State)

Eight representative queries were tested live. Pattern observed:

Key takeaways:

1. Branded queries are locked down. Good — but these are low-value (high-intent buyers already know the brand).
2. Definitional queries = the biggest miss. TBOF has no content answering "what is A2 bilona ghee / how is it made / is it worth it" on the US store. Competitors own this surface, which is precisely what AI Overviews and ChatGPT retrieve and cite for top-of-funnel ghee questions.
3. Comparison queries are owned by third parties. Authentic Urban, Heroveda, Shahji, and Raghas all publish "Two Brothers vs X" pages that mention TBOF but frame it on their terms. TBOF has zero "vs" content on the US store to counter-narrate. Per the Princeton GEO study, comparison articles are the single most-cited AI content type (~33%).
4. The India store (twobrothersindiashop.com) is doing the visibility lifting on "worth the price" via a strong blog post — the US store should have its own equivalent.

4. Content Extractability Audit

Critical content defects:

• Homepage has no <h1>. Every other key page (about, contact, policies, blog) has one. The homepage is the canonical entity-definition page for AI — its absence is a direct extractability loss.
• The FAQ page is effectively empty (312 visible chars; the "questions" are theme placeholders [[ title ]]). It returns no citable Q&A and carries no FAQPage schema. This is a wasted, high-value AI surface.
• "Two Brothers India Farms" naming on the US store. The homepage copy says "No Tariffs. Just Tradition — Two Brothers India Farms," "What Makes Two Brothers India Farms Different?," "Why Americans Are Choosing Two Brothers India Farms." This name does not match the legal entity (TBOF FOODS PVT LTD), the About page ("Two Brothers Organic Farms"), or the domain brand ("USA Two Brothers Farms"). AI/Google Knowledge Graph resolve on consistent entity names. This fragmentation is a measurable citation drag and should be unified to "Two Brothers Organic Farms" (or "Two Brothers Organic Farms USA") everywhere on the US store.
• No comparison/guide content on the US store (only 2 "why ghee" pages of 83 total pages). The highest-cited AI content type is absent.

5. Priority Recommendations (Roadmap)

Ordered by impact-to-effort. Each maps to a pillar: Structure / Authority / Presence.

P0 — Fix entity & structural integrity (do first; low effort, high impact)

1. Unify the brand name on the US store. Replace "Two Brothers India Farms" → "Two Brothers Organic Farms" (or "Two Brothers Organic Farms USA") across homepage hero, section headings, meta title/description, and Organization/LocalBusiness schema name field. Keep "USA" as a differentiator in copy, not as the entity name. (Structure + Authority)
2. Add an <h1> to the homepage with a direct definitional answer: e.g. "Two Brothers Organic Farms — A2 Bilona Ghee & Regenerative Organic Foods, Shipped Tariff-Free from the USA." Lead the page with a 40–60 word answer block defining the brand and hero product. (Structure)
3. Repair the FAQ page. Replace [[ title ]] placeholders with 8–12 real, natural-language Q&As (shipping, tariffs, A2 vs regular ghee, shelf life, subscriptions, authenticity, returns) AND add FAQPage JSON-LD. This becomes an instant AI-citation surface. (Structure + Authority)

P1 — Build the definitional + comparison content layer (biggest visibility win)

1. Publish a definitive "What is A2 Bilona Ghee?" guide (US store, /pages/what-is-a2-bilona-ghee) with: 40–60 word definition block up top, a step-by-step "how it's made (curd → churn → simmer)" block, a comparison table (A2 Bilona vs regular/cream ghee vs butter), and FAQPage schema. This directly targets the query TBOF currently loses to 6 competitors. (Authority + Structure)
2. Publish "Two Brothers vs [competitor]" comparison pages for the 3 closest US-relevant analogues (Anveshan, Barosi, and one commodity brand e.g. Amul/24 Mantra). Balanced, structured, with a comparison table and a clear "why TBOF" differentiator section. Competitors already write these about TBOF — counter-narrate on TBOF's own domain. (Authority + Presence)
3. "Is A2 ghee worth the price?" justification page (mirror the strong India-store post) with cost-per-use math, 30L-milk-to-1L-ghee context, and ECOCERT/certification proof. Add FAQPage schema. (Authority)
4. Cross-link blog content to PDPs. The 16k-char "why ghee" page and food-health blog posts should link to the relevant ghee/flour/oil PDPs so AI retrieval of the guide also surfaces product pages. (Presence)

P2 — Machine-readable completeness (small, high-leverage)

1. Add /pricing.md (or /pricing.txt) — a structured, parseable price list (ghee tiers $29–$289, flours, oils, bundles, free-$99-shipping threshold). AI buying agents and Perplexity-style engines reward transparent, readable pricing; today it 404s. Link it from llms.txt/agents.md and the sitemap. (Presence + Structure)
2. Unblock /policies/ in robots.txt (or add an Allow: /policies/ exception scoped to AI bots). Shipping/refund/return promises are trust signals AI engines extract; right now they're Disallowed even though the pages return 200. Add ShippingDelivery/Offer/MerchantReturnPolicy schema to those pages while at it. (Authority)
3. Add ItemList schema to collection pages (/collections/ghee, etc.) and consider FAQPage on blog posts that contain Q&A. Currently collections expose only LocalBusiness+BreadcrumbList. (Structure)
4. Refresh "last updated" dates on the core guide/FAQ pages and surface author attribution (founder names + credentials) on the definitional content — E-E-A-T signals AI engines weight heavily. (Authority)

P3 — Presence (ongoing, lower effort here)

1. Wikipedia / entity accuracy: Ensure the brand's Wikidata/Wikipedia entry (if any) and Google Business Profile use the consistent "Two Brothers Organic Farms" name and link the US store. Third-party mentions drive ~7.8% of ChatGPT citations.
2. Maintain the existing agentic endpoints (llms.txt, UCP) — these are a genuine differentiator; keep them current as the catalog changes.
3. Get the US Amazon storefront review velocity up (currently relies on India's 1,796 reviews). US-specific ratings feed both Amazon A9 and AI shopping answers.

6. What NOT To Do (per Google's AI guidance)

• Do not write separate "for AI" pages — the guides above serve humans first; good structure is a bonus, not a gimmick.
• Do not chunk the homepage into AI-bait fragments — add one H1 + one answer block; keep normal prose.
• Do not block AI bots — already correct; keep GPTBot/PerplexityBot/ClaudeBot/Google-Extended allowed.
• Do not keyword-stuff ghee/AYV phrases — it actively reduces AI visibility (−10% per Princeton GEO).
• Do not gate the guides behind email/JS — keep them public and server-rendered so both Search and agents read them.

7. Expected Impact

The store already wins branded and pricing queries. The prize is the definitional and comparison layers (P1), which are where competitors currently own the AI narrative for TBOF's own hero category.

8. Sources & Verification

• Live robots.txt: https://twobrothersfood.com/robots.txt
• Agentic endpoints: https://twobrothersfood.com/llms.txt · https://twobrothersfood.com/agents.md · https://twobrothersfood.com/.well-known/ucp
• Pricing file (404): https://twobrothersfood.com/pricing.md
• Schema (raw HTML JSON-LD): homepage, https://twobrothersfood.com/products/amorearth-desi-cow-a2-ghee, https://twobrothersfood.com/pages/faqs, https://twobrothersfood.com/blogs/food-health/wood-pressed-oil-vs-refined-oil, https://twobrothersfood.com/collections/ghee
• Sitemaps: https://twobrothersfood.com/sitemap.xml (+ products/pages/collections/blogs/agentic submaps)
• Brand-name counts + H1 absence: verified in rendered homepage/about/contact HTML (2026-08-26)
• Visibility queries: live web search on "what is A2 bilona ghee," "best A2 ghee USA two brothers," "two brothers vs anveshan," "is A2 bilona ghee worth the price," "two brothers organic farms ghee price USA" (2026-08-26)
• Business context: business-analysis.md — 61 SKUs, $6–$289, 69.2% phone audience, ₹110 Cr Series B, A2 Bilona hero, ECOCERT, diaspora US positioning
• Competitive landscape: Anveshan, Barosi, Delight Farms, Ritvaam, Authentic Urban (per business-analysis.md §19)

No follower counts, ad-spend figures, or revenue numbers were inferred or fabricated; where the upstream business analysis left values blank (Instagram/FB login walls), they remain so. Schema verification used raw HTML extraction, not cleaned-text tools, to avoid false "no schema" findings.