AEO / GEO Audit
Rynox Gear — AI Search Optimization (AEO / GEO / LLMO) Audit
Domain: rynoxgear.com | Prepared: 2026-08-27 | Method: live crawl of rendered HTML + JSON-LD, robots.txt/llms.txt/agents.md/UCP endpoints, and proxy AI-visibility checks via web_search across key queries. Every finding is tagged [Verified] (directly observed) or [Inferred]. Business context from business-analysis.md (27-section e-commerce analysis) is used throughout.
Executive Summary
Rynox Gear is unusually well-positioned for the AI-search era on the machine-readable / agentic front — but weak on the content extractability / authority front that actually drives citations in ChatGPT, Perplexity, Gemini and Google AI Overviews.
What's already strong (rare for a Shopify D2C brand):
/llms.txt,/agents.md,/robots.txt,/.well-known/ucpand/api/ucp/mcpall return HTTP 200 and are functional (UCP discovery and MCPtools/listreturn real schema, not errors). An AI agent can already read the catalog and even begin a buyer-approved checkout. No AI crawler is blocked — GPTBot, PerplexityBot, ClaudeBot, Google-Extended and Bingbot are all implicitly allowed (only/admin,/cart/,/checkout,/accountetc. are disallowed).- Product pages carry real
Product+OrganizationJSON-LD, includingaggregateRatingandBreadcrumbList.
What's blocking AI citation today:
- No extractable answer content. The blog is frozen (last post Mar 2022; ~13 visible posts only), there are no FAQ pages, no comparison articles, no "how to choose" guides. AI engines cite comparison articles (~33% of citations), definitive guides (~15%) and original data (~12%) — Rynox has none of these in a citable form.
- Third-party sources dominate the brand's narrative. In proxy searches for "rynox gear review", "rynox vs viaterra", and founder/history queries, Rynox is out-cited by Team-BHP, BikeWale, RyderPlanet, Tracxn, and a founder-story blog. Per the Princeton GEO study, brands are 6.5x more likely to be cited via third parties — but Rynox has far more third-party coverage than it controls on its own site, and almost none of that coverage is structured for easy extraction.
- No FAQ/HowTo/Article schema, no
pricing.md. PDPs haveProduct/aggregateRating/BreadcrumbListbut notFAQPageorHowTo(e.g. "how to fit armour", "how to size a riding jacket")./pricing.mdand/pricing.txtboth 404 — there is no machine-readable price ladder for agents. - Thin entity presence off-domain. No Wikipedia page; social profiles exist (151K IG, 40K FB, 10.7K LinkedIn) but are not yet leveraged for entity signals.
Bottom line: Rynox has solved the plumbing (agents can reach it) but not the substance (nothing worth citing). The single highest-ROI move is to revive the content engine with extractable, stat-backed, FAQ/comparison structures — which also fixes the stale-blog weakness already flagged in the business analysis.
1. AI Bot Access & Machine-Readable Files [Verified]
| File / Endpoint | Status | Finding |
|---|---|---|
robots.txt | HTTP 200 | No Disallow for GPTBot, PerplexityBot, ClaudeBot, Google-Extended, Bingbot, CCBot. Only transactional paths blocked (/admin, /cart/, /checkout, /account, /orders). AI crawlers can read the store. ✅ |
llms.txt | HTTP 200 | Present but generic boilerplate. Shopify/UCP auto-generated text referencing the Shop skill and UCP flow — not Rynox-specific product/brand context. It does mirror /agents.md. |
agents.md | HTTP 200 | Same auto-generated UCP/shopping-assistant boilerplate as llms.txt. No Rynox-specific answer content. |
.well-known/ucp | HTTP 200 | Real, functional UCP discovery (version 2026-04-08, services listed). Not an error stub. |
api/ucp/mcp | HTTP 200 | Real, functional MCP endpoint — tools/list returns actual shopping tools (get_checkout, etc.). Agents can transact (buyer-approved). |
pricing.md | HTTP 404 | Gap — no machine-readable price ladder for AI agents. |
pricing.txt | HTTP 404 | Gap — same. |
| Wikipedia page | none found | Gap — no English Wikipedia article for Rynox Gear. |
Interpretation: Rynox is ahead of ~95% of Shopify stores on agentic readiness (real UCP/MCP). But llms.txt/agents.md are not doing the citable-content job the skill describes — they're checkout plumbing, not a context file that would help a non-transacting AI answer "what is the best Rynox jacket for touring?" Add a genuine Rynox section to llms.txt (or a separate context page) describing the brand, hero products, and links to key pages, distinct from the auto-generated UCP block.
2. Schema Markup (verified via raw HTML, not text-extraction) [Verified]
JSON-LD was checked with curl + script-tag-preserved parsing (the reliable method — markdown/text extraction strips <script> blocks and would falsely report "zero schema").
| Page | JSON-LD present | Types | Notes |
|---|---|---|---|
| Homepage | 5 blocks | Organization, WebSite, 3× Product | Good. Featured products exposed. |
| Product (PDP) | 2 blocks | Organization, Product | Product has name, image, description, sku, brand, offers, aggregateRating (real Judge.me review count). ✅ |
| PDP | separate | BreadcrumbList | Present (good for AI category understanding). |
| About / policies | 0 blocks | — | No Article/FAQPage/HowTo on any content/info page. |
Gaps:
- No
FAQPageschema anywhere — despite riding gear being a high-FAQ category (sizing, CE levels, armour types, care). This is the highest-value missing schema for AI extraction. - No
HowToschema — "how to choose a riding jacket", "how to fit D3O armour", "how to pack a tank bag" are exactly the process queries AI Overviews answer. - No
Article/BlogPostingon the (stale) blog — blog posts aren't marked up for author/date extraction, weakening freshness signals. - PDPs have
aggregateRatingbut the homepage/collections don't surface aggregate ratings for category-level AI answers.
3. Content Extractability [Verified + Inferred]
| Check | Result |
|---|---|
| Clear definition in first paragraph (homepage)? | Partial — tagline only; no "What is Rynox" answer block. |
| Self-contained 40–60 word answer blocks? | No — no Q&A/definition blocks anywhere. |
| Statistics with sources? | No on-site. (Business analysis has them, site doesn't.) |
| Comparison tables for "X vs Y"? | No — biggest missed opportunity vs Viaterra/Solace/DSG. |
| FAQ section with natural-language questions? | No. |
| Expert attribution (named authors)? | No — blog is anonymous, frozen since 2022. |
| Recently updated (within 6 months)? | No — blog last updated Mar 2022 (~4 yrs stale). |
| Heading structure matches query patterns? | Partial — category headings exist; no question-style H2s. |
Extractability verdict: Low. The site sells well but explains little. An AI answering "what riding jacket should I buy for ₹10k touring in India?" has almost nothing Rynox-authored to cite — it falls back to third-party reviews (Team-BHP, BikeWale), which is why Rynox appears as a subject in AI answers more often than a source.
4. AI Visibility — Proxy Citation Check [Verified via web_search]
Proxy tests (web_search approximates the organic/AI-retrieval corpus; not a live ChatGPT/Perplexity call, but indicative of what's indexed and citable):
| Query | Rynox Cited? | Who/What Appears | Implication |
|---|---|---|---|
| "best riding jacket india ce certified" | No (ranked below YouTube/guides) | YouTube, Hygena, Bikester, Shortgears | Category-intent query → Rynox absent; generic guides win. |
| "rynox gear review riding jacket" | Yes (own site #1) + 3rd-party reviews (Team-BHP #2, RyderPlanet #3, BikeWale #5) | Brand + reviewers | Brand-name queries do well; third parties out-cite on opinion. |
| "motorcycle tank bag india rynox vs viaterra" | Partial (homepage #1) but Viaterra-specific reviews dominate results | Rynox home, Viaterra reviews, brahmandtour | "vs" comparison → Rynox has no own comparison page; loses depth. |
| "rynox gear founded history mumbai manufacturer" | Yes — About Us #2, Tracxn #1, founder blog #5 | Tracxn, About Us, werisebyliftingothers | Founder/entity queries well covered (good E-E-A-T source). |
Pattern: Rynox wins branded and entity queries but loses category and comparison queries to guides and competitors. This is the classic "strong domain, no answer content" gap. Fixing it (Section 6) would lift AI Overviews and ChatGPT/Perplexity citations for the high-intent, high-value queries that actually drive sales.
5. Topical Clusters & Query Fan-Out [Inferred from catalog + competitors]
Google AI search fans out from a single query into related sub-queries. Rynox should own these clusters (all currently thin or absent on-site):
- "Best riding jacket for [touring / commuting / off-road / under ₹10k]" → no buying-guide content.
- "Rynox vs [Viaterra / Solace / DSG / Reise]" → no comparison pages; competitors win by default.
- "What is CE Class AA / D3O / Rheon armour?" → no explainer; Rynox sells certification but doesn't explain it (a citable authority play).
- "How to choose / size / care for riding gear" → no HowTo content.
- "Rynox warranty / Wingman insurance / Crash Replacement Program" → trust programmes exist but aren't structured as extractable FAQ (high citation value — unique differentiators).
- "Motorcycle luggage [tank bag / saddlebag / tail bag] buying guide" → Rynox is a luggage leader but has no guide; Viaterra owns the comparison narrative.
Action: build one pillar page per cluster with definition blocks, comparison tables, FAQ schema, and cited stats — this simultaneously satisfies Google's fan-out and non-Google AI engines' extractable-structure preference.
6. Prioritized Recommendations
Ordered by AI-visibility impact × feasibility (feasibility informed by business-analysis.md).
| # | Recommendation | Pillar | Impact | Effort | Why |
|---|---|---|---|---|---|
| 1 | Revive the content engine with extractable buying-guides, "vs" comparisons, and explainers (CE/D3O/Rheon, jacket sizing, luggage types). Add FAQPage + HowTo schema. | Structure + Authority | High | Low–Med | Directly fills the citation gap; also reverses the 2022 blog freeze (already flagged in business analysis). Comparison articles = ~33% of AI citations. |
| 2 | Add FAQ sections + FAQPage schema to every PDP and key info pages (sizing, CE levels, Wingman/Crash Replacement, shipping/returns). | Structure | High | Low | Cheap, high-yield; answers the natural-language questions AI extracts. |
| 3 | Create /pricing.md (machine-readable price ladder by category: ₹200–₹18,900, median ₹5,200) and link from sitemap + footer. | Presence + Agentic | Med | Low | Agents comparing products programmatically need parseable pricing; currently 404. |
| 4 | Make llms.txt Rynox-specific — add a real brand/hero-product/link context block (distinct from the auto-generated UCP plumbing). | Presence | Med | Low | Today it's generic boilerplate; a genuine context block helps non-transacting AI answer brand/category questions. |
| 5 | Build "Rynox vs [competitor]" comparison pages for Viaterra, Solace, DSG, Reise (luggage + jackets). | Authority + Structure | High | Med | Reclaims the comparison-query narrative currently owned by third parties. |
| 6 | Strengthen off-domain entity presence — pursue a Wikipedia mention, get featured in rider-media roundups (Team-BHP, BikeWale, OTOCapital), answer Quora/Reddit riding-gear questions authentically. | Presence | Med | Med | 6.5x citation lift via third parties; Rynox already has strong but unstructured third-party coverage to build on. |
| 7 | Add author/expert attribution + "Last updated" dates to all new content (founder/racing credentials, e.g. #ProjectRAID, Dakar). | Authority | Med | Low | Named expertise + freshness are top GEO ranking factors (+25–30% citation). |
| 8 | Surface aggregate ratings on collections/homepage (not just PDP) so category-level AI answers can cite Rynox review volume. | Structure | Low–Med | Low | Currently only PDPs expose aggregateRating. |
| 9 | Keep UCP/MCP live and monitor — already best-in-class; ensure it stays functional through theme/app changes (REBUY app in use). | Agentic | Maintain | Low | Future-proofs agentic/buy-for-me commerce. |
Do NOT: write separate "AI-only" content (Google's spam policy), keyword-stuff, gate the guides behind login, or block AI crawlers. Keep all new content people-first with normal headings (Google's explicit guidance).
7. Monitoring (DIY, no paid tool required)
Monthly, for the top 15–20 queries in Section 4/5:
- Run each through ChatGPT (with search), Perplexity, and Google — record: AI Overview present? Rynox cited? Which page? Competitors cited?
- Track brand citation rate and share of AI voice in a sheet.
- Re-run Rich Results Test on PDPs after adding FAQPage/HowTo schema.
- Watch for a Wikipedia article and new third-party reviews to cite.
Paid cross-platform tools if budget allows: Otterly AI, Peec AI, LLMrefs (ChatGPT/Perplexity/Gemini/Claude/AO tracking).
8. Citations (sources used in this audit)
- https://rynoxgear.com/ — storefront, rendered HTML + JSON-LD, robots.txt, llms.txt, agents.md, /.well-known/ucp, /api/ucp/mcp (all directly crawled 2026-08-27)
- https://rynoxgear.com/products.json — 103 products / 818 variants (catalog for topical clusters)
- https://rynoxgear.com/blogs/news — blog confirmed frozen (last post Mar 2022; ~13 visible posts)
- https://rynoxgear.com/pages/about-us — founder/entity narrative (ranks for founder queries)
- https://tracxn.com/d/companies/rynox-gear/ — founded 2011/2012, unfunded, legal entities (entity queries)
- https://www.team-bhp.com/news/motorcycle-riding-gear-review-rynox-cypher-jacket-viaterra-gloves — third-party review out-citing Rynox on opinion queries
- https://www.bikewale.com/expert-reviews/rynox-dune-neo-trail-off-road-jacket-and-pants-long-term-review-introduction/ — third-party long-term review
- https://www.werisebyliftingothers.in/2024/10/journey-of-rahul-katwa-founder-and-ceo.html — founder story (E-E-A-T source)
- https://www.otocapital.in/article/top-5-motorcycle-riding-gear-brands-in-india — competitor landscape (Rynox, DSG, Solace)
- Princeton GEO study (KDD 2024) — citation-method boost figures (+40% cite sources, +37% statistics, comparison ~33% of citations, 6.5x third-party citation lift)
- Google AI features optimization guide (developers.google.com/search/docs/fundamentals/ai-optimization-guide) — "no special markup required", don't chunk/write-for-AI stance
- business-analysis.md (this workspace) — 27-section baseline: 103 SKU, ₹200–₹18,900 pricing, competitors (Viaterra/Solace/DSG/Reise), founder/entity facts, stale-blog gap
Confidence legend: [Verified] directly observed via crawl/JSON-LD/endpoint. [Inferred] reasoned from catalog, competitor set, and GEO research patterns. Web_search proxy checks approximate the indexed/retrieval corpus and are directional, not live ChatGPT/Perplexity output.