sundried.com — AI Site Grade

Analysis

Sundried's AI-Visibility Audit

The site has a fully functional llms.txt and agents.md plus a live Universal Commerce Protocol endpoint — making it one of the most technically AI-ready Shopify stores found — yet its own brand story, sustainability credentials, and product catalog are almost invisible to the content layer AI models actually read.

Crawler Access

All major AI bots (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, Bytespider, Applebot-Extended) receive 200 status with full HTML from https://www.sundried.com. The non-www variant (https://sundried.com) triggers a Cloudflare JS challenge wall that blocks plain GET requests, but the www canonical resolves cleanly. The robots.txt is a standard Shopify-generated file with no AI-bot-specific directives — no Disallow: / for any crawler. The llms.txt exists and returns 200 with agent instructions directing shopping agents to use the Shop skill and the UCP protocol. The agents.md file and /.well-known/ucp endpoint are both live, implementing the Universal Commerce Protocol for agent-driven checkout. This is an unusually sophisticated AI-commerce setup for a mid-market activewear brand.

Cold-Knowledge Gap

The LLM knows Sundried as a UK-based sustainable activewear brand, certified B Corporation, using recycled materials and coffee grounds, with named products like Tempest, Pioneer, and Apex. The actual site tells a different story. The homepage contains 87 words of visible text — no mention of B Corp, no sustainability manifesto, no brand origin story. The /pages/about-us and /pages/sustainability URLs both return 404. The product pages mention "Eco Tech range" and "recycled coffee grounds" but only in passing. The Organization schema on every page has empty strings for four of eight sameAs fields. The B Corp certification the model cites as a "well-known fact" is entirely absent from the site — no badge, no page, no mention anywhere in the fetched content. Either the certification lapsed or was never formally held, but the model's prior is stronger on sustainability positioning than the site itself delivers.

Content & Schema Posture

The site runs on Shopify with a blog-heavy content strategy that is almost entirely third-party gear reviews and training guides — Coros watches, Hoka shoes, Garmin bike computers, Bollé eyewear. These blog posts are well-written, long-form, and use Article schema with author markup. However, the blog index page lists 32 pages of posts, and several posts linked from the homepage (e.g., the Coros Pace 4 review) return 404. The blog is a content marketing engine for other brands' products, not Sundried's own. Product pages have thin content (~94 words) and lack Product schema — only BreadcrumbList and Organization are present. No Product, Offer, AggregateRating, or Review structured data exists on any product page checked. The WebSite schema includes a SearchAction target, which is correct.

External Signals

Web searches for Sundried across review sites, Reddit, and B Corp directories returned zero results — no Trustpilot page, no Reddit threads, no press coverage surfaced. The DNS TXT records show Google and Facebook verification plus Mailchimp/Google email, but no B Corp verification token. The Wayback Machine shows a snapshot from May 2026, confirming the site is live and has been for some time, yet the brand has essentially no discoverable off-domain footprint in current search indices.

Findings

1. Brand story and sustainability credentials absent from site content High

   The homepage contains only 87 words of visible text with no mention of B Corp, sustainability, or brand origin. The /pages/about-us and /pages/sustainability URLs return 404. The B Corp certification cited by LLMs is entirely absent from the site.

   What to change: Add a dedicated sustainability page and about page with B Corp badge, brand story, and sustainability claims. Enrich homepage copy with key brand differentiators.

2. Product pages lack Product structured data High

   Product pages have thin content (~94 words) and no Product, Offer, AggregateRating, or Review schema. Only BreadcrumbList and Organization are present.

   What to change: Add Product, Offer, and AggregateRating structured data to all product pages.

3. Blog posts linked from homepage return 404 Medium

   Several blog posts linked from the homepage, such as the Coros Pace 4 review, return 404 errors. The blog index lists 32 pages of posts, but content is missing.

   What to change: Fix or redirect broken blog post URLs. Ensure all linked content resolves to 200.

4. Organization schema has empty sameAs fields Medium

   The Organization schema on every page has empty strings for four of eight sameAs fields, reducing credibility signals to AI crawlers.

   What to change: Populate all sameAs fields in Organization schema with actual social media and review profile URLs.

5. No discoverable external signals from reviews, social, or press High

   Web searches for Sundried across Trustpilot, Reddit, B Corp directory, and general reviews returned zero results. The brand has no off-domain footprint in current search indices.

   What to change: Build external signals by encouraging reviews on Trustpilot, engaging in relevant communities, and securing press coverage. Verify B Corp status and add verification token to DNS.

6. Product pages have thin visible text content Medium

   Product pages contain only about 94 words of visible text, limiting the information available to AI crawlers for indexing and understanding.

   What to change: Expand product descriptions with detailed features, materials, sizing, and sustainability attributes.

7. Non-www domain triggers Cloudflare JS challenge Low

   The non-www variant (https://sundried.com) returns a Cloudflare JS challenge that blocks plain GET requests, potentially confusing some crawlers.

   What to change: Redirect the non-www domain to the www canonical at the DNS or server level to avoid crawler confusion.

8. Blog content focuses on third-party products, not Sundried's own Medium

   The blog is dominated by gear reviews and training guides for other brands (Coros, Hoka, Garmin, Bollé), with little content about Sundried's own products or brand story.

   What to change: Balance the blog with original content about Sundried's products, sustainability initiatives, and brand values.

What's working

• llms.txt file is live and well-structured — The site hosts a functional llms.txt that returns 200 and provides agent instructions for shopping and UCP usage, making it AI-commerce ready.
• agents.md and Universal Commerce Protocol endpoint are live — The agents.md file and /.well-known/ucp endpoint are both live, enabling agent-driven checkout via the Universal Commerce Protocol.
• All major AI bots receive 200 with full HTML — All tested AI bots (GPTBot, ClaudeBot, etc.) receive 200 status with full HTML from the www domain, with no disallow directives in robots.txt.
• Sitemap is accessible and contains 80 URLs — The sitemap returns 200 and lists 80 URLs, providing a clear crawl path for search engines and AI crawlers.
• Blog posts use Article schema with author markup — Blog posts implement Article structured data with author information, aiding AI crawlers in understanding content authorship.
• WebSite schema includes SearchAction target — The WebSite schema correctly includes a SearchAction target, enabling AI agents to perform site search.
• Protein calculator page provides interactive tool content — The protein calculator page offers useful interactive content with 169 words of text, adding value for users and crawlers.
• Product collection page has substantial text content — The /collections/all page contains 1152 words of text, providing rich content for indexing product categories.