Original homepage scans: July 2026. Interpretation and limitations reviewed September 27, 2026.
This article preserves the original reported homepage audit scores for 17 San Diego County brands. The automated audit checked 21 technical and content signals, including structured data, entity information, FAQ markup, content depth and SEO basics. It did not ask AI engines to recommend the brands or measure how frequently their websites were cited.
Historical technical-readiness scores
- Blenders Eyewear — 91
- Pura Vida Bracelets — 91
- Suja Juice — 88
- Reef — 77
- Firewire Surfboards — 74
- RAEN — 74
- Linksoul — 74
- Sun Bum — 69
- Salty Crew — 68
- Rusty — 66
- Nixon — 62
- Matuse — 58
- Stone Brewing — 50
- Electra Bikes — 46
- JuneShine — 43
- Cutwater Spirits — 33
- TaylorMade — 11
Scores are out of 100 and reflect the original July report, not a fresh audit, product quality, brand reputation or verified AI-search performance.
What the original observations can support
The original report detected differences in homepage markup and content available to its scanner. It reported no detectable homepage FAQ schema in the listed sample. That does not establish whether useful FAQs existed elsewhere, whether another crawler could render more content, or whether an engine used the brand’s website in an answer.
This local snapshot followed the 16-brand outdoor readiness snapshot. Comparing the two requires the same scanner version and method; neither is a ranking of AI recommendations.
What this measures—and what it does not
This is a technical website-readiness snapshot, not a test of how often ChatGPT, Perplexity or Google recommends a business. The score reflects the audit tool’s chosen checks and weights. Those weights have not been validated as predictors of rankings, citations or leads. A homepage scan can miss useful content on other pages or content loaded after JavaScript runs.
Missing FAQ markup, an llms.txt file or a particular schema type does not make a business invisible to AI. An unsuccessful scan also does not prove that an official search crawler is blocked. Different crawlers and browsing tools can receive different responses.
How to measure actual AI visibility
- Define the buyer questions, location, search-enabled engine and test date before running prompts.
- Repeat the same questions in independent sessions, without telling the engine to recommend your business.
- Save the complete answers and citation URLs. Count business mentions separately from links to the business’s own site.
- Check factual accuracy, especially prices, service areas and delivery terms. Record failed or inaccessible runs.
- Compare visits and qualified inquiries separately; a citation is not a sale.
This edition does not contain that prompt-based dataset. Its technical scores must not be presented as recommendation frequency or market share.
Sources and corrections
Google’s AI search guidance explains that ordinary search eligibility and useful content matter; no special AI text file or schema is required. llms.txt is a proposal for supplementary site information, not a ranking guarantee.
Clarified September 27, 2026: the original URLs and historical score lists are retained, but claims that low scores prove AI invisibility or that markup guarantees citations have been removed. The original article did not publish complete raw scan responses or a versioned scoring specification, so these historical numbers cannot be independently reproduced from this page alone. They have not been rerun or independently verified in this update.
Use the free technical-readiness audit to identify items for human review. For help prioritizing content and technical changes, explore Kanyon Studio’s search optimization service or send a project brief.