Skip to main content
案例分析

【AI Visibility Case Study】Strong SEO, Zero AI Presence — 47/100

Published on March 24, 2026

Case Overview

This case examines a Taiwan-based medical device contract manufacturing and OEM services company. On paper, its website looked solid — page titles and meta descriptions were properly configured, and the technical SEO score reached an impressive 92 out of 100, placing it well ahead of many industry peers. Yet when we queried leading AI platforms using the exact search terms a real buyer would use — phrases like "medical device contract manufacturer" or "Medical device OEM supplier" — the company received zero mentions across every industry-level query. Not once was it recommended.

This contradiction is the heart of the case: strong traditional SEO performance does not translate into AI visibility. The comprehensive health check score came in at 47/100, with an AI visibility potential rating of "Moderate." The good news is that the problem structure is clear, and there is a defined optimization path forward.

Score Breakdown

The health check evaluated this medical device contract manufacturer across three dimensions. The wide gap between scores reveals exactly where the problem lies.

Evaluation DimensionScoreStatus
AI Brand Mention Rate40 / 100⚠️ Needs Improvement
GEO Technical Audit33 / 100🔴 High Risk
Website Performance (PageSpeed)71 / 100🟡 Acceptable
Overall Score47 / 100⚠️ Moderate Potential

The primary bottleneck is unmistakable: the GEO technical audit score of just 33 is the main drag on overall performance. The near-complete absence of structured data means AI systems cannot effectively interpret or cite the company's professional positioning — creating a situation where the website exists but AI simply cannot understand it.

AI Search Visibility Testing

The real test of AI visibility is whether AI platforms will proactively recommend your business when a buyer asks. We submitted queries to three major AI platforms — Claude, ChatGPT, and Gemini — covering both brand-specific and industry-scenario queries, for a total of six tests. The results were strikingly consistent: partial performance on brand queries, and a complete shutout on industry queries.

Claude

The brand query returned a △ Vague Mention — Claude did not clearly affirm the company's expertise or service scope. The response was cautious and lacked specific detail. The industry query (e.g., "recommended medical device contract manufacturers in Taiwan") returned ✗ Not Mentioned. This means that even if the brand name exists somewhere on the edge of Claude's knowledge base, it does not appear in recommendation lists when buyers are making real purchasing decisions.

ChatGPT

The brand query received a ✓ Positive Mention — the strongest result across all three platforms. ChatGPT recognized the brand and provided a basic positive description. However, the industry query again returned ✗ Not Mentioned. When buyers search by service type rather than brand name, the company disappears from the results entirely. This is a critical insight: brand recognition is not the same as scenario-level AI visibility.

Gemini

Results mirrored ChatGPT — ✓ Positive Mention for the brand query, ✗ Not Mentioned for the industry query. Gemini has basic awareness of the brand but cannot surface the company in procurement-oriented queries such as "medical device OEM supplier."

In summary: 3 out of 6 queries mentioned the brand, but 0 out of 6 industry-scenario queries resulted in a recommendation. This is precisely the fundamental difference between AI visibility optimization and traditional SEO. You need AI to find you before buyers even know your name.

Competitive Landscape

When AI platforms are asked about medical device contract manufacturers, the companies they tend to recommend are those that have invested heavily in structured content, strong E-E-A-T signals, and detailed technical documentation in English. Many are internationally recognized contract manufacturing groups, or companies that have publicly disclosed FDA 510(k) clearance, ISO 13485 certification, and comprehensive supply chain information.

These companies continuously reinforce AI systems' understanding of their expertise through whitepapers, certification guides, process documentation, and long-form technical content. By contrast, the company in this case — despite having genuine manufacturing capabilities — lacks structured content assets, leaving it at a competitive disadvantage in the AI recommendation ecosystem. As AI-driven search continues to grow in adoption, this gap will translate into measurable lost inquiries if left unaddressed.

GEO Technical Audit

GEO (Generative Engine Optimization) technical fundamentals are the underlying architecture of AI visibility. The audit assessed 9 key technical indicators. The company passed only 4, for a pass rate of 4/9 (approximately 44%) and a technical audit score of just 33/100.

Technical ItemStatus
Schema JSON-LD Structured Data✗ Not Configured
XML Sitemap✗ Not Configured
Meta Description✓ Configured
OG Tags (Social Preview Tags)✗ Not Configured
Canonical URL✗ Not Configured
HTTP/2 Protocol✗ Not Enabled
Title Tag✓ Configured
H1 Tag✗ Not Configured
PageSpeed Overall✓ 71 (Acceptable)

The most critical gap is the complete absence of Schema JSON-LD structured data. For a medical device contract manufacturer, Organization Schema, Service Schema, and Certification markup are the core signals that tell AI systems what a company does and what qualifications it holds. Without them, the company's ISO certifications, manufacturing credentials, and core competencies are effectively invisible to AI.

The missing H1 tag means page topic signals are ambiguous. The absence of a sitemap reduces AI crawler indexing efficiency. Together, these issues compound the AI visibility loss significantly.

Website Performance

Website performance directly affects how willing AI crawlers are to fetch and index content. The overall PageSpeed score came in at 71/100, but digging into the sub-scores reveals a stark internal contradiction: Performance was only 50/100, while the technical SEO score reached 92/100 — a gap of 42 points.

A Performance score of 50 indicates slow page load times, likely caused by uncompressed image assets, render-blocking CSS or JavaScript, or a lack of lazy loading. For AI crawlers conducting large-scale indexing, slow-loading pages are routinely deprioritized, directly reducing the chances that content will be cited. The target should be to bring the Performance score above 75, which would meaningfully improve content accessibility across AI platforms and support stronger overall AI visibility.

Expert Diagnostic Recommendations

Based on the audit data, we identified three optimization priorities with the highest leverage effect. The following is a diagnostic summary; a complete optimization plan requires a customized assessment based on the company's specific situation.

1. Close the Performance and Crawler Accessibility Gap First

A Performance score of 50 is the most immediate technical factor suppressing AI indexing efficiency. The symptoms point to insufficient image optimization and a weak asset loading strategy. Notably, fixing this issue will produce faster AI visibility improvements in the short term than adding new content — because content on pages that AI crawlers cannot fully load will never be cited, regardless of quality.

2. Structured Data Absence Prevents AI from Understanding Core Expertise

The complete lack of Schema JSON-LD is the primary reason the technical audit score is so low. The deeper issue is that in a highly specialized B2B field like medical device contract manufacturing, AI systems require explicit structured signals to classify a company as a "credible manufacturer" rather than a generic supplier. Without Certification, Service, and Organization Schema markup, the company's ISO certifications, regulatory qualifications, and manufacturing capabilities are invisible to AI systems — and therefore absent from AI-generated recommendations.

3. Bridge the Systematic Gap Between Brand and Industry Mentions

Three platforms mentioned the brand on direct brand queries, yet zero industry queries resulted in a recommendation. This gap reflects a systematic absence of professional scenario-driven content. Medical device buyers in the early stages of procurement do not typically search by vendor name — they search by manufacturing process, certification type, or application category. The company's current content assets do not cover these pre-purchase query scenarios, leaving a critical blind spot in its AI visibility footprint.

AI Search Trends in Medical Device Contract Manufacturing

Procurement decisions in the medical device contract manufacturing and OEM sector are inherently trust-driven, long-cycle, and multi-layered in validation. A procurement director from a European or American medical brand evaluating Taiwan-based contract manufacturers typically navigates a months-long supplier selection process: initial market research, capability assessment, certification verification, factory audits, sample testing, and finally pricing and contracts.

At the very front end of this process — the market research and initial shortlisting stage — AI search is playing an increasingly decisive role. When procurement teams ask ChatGPT or Gemini questions like "ISO 13485 certified medical device contract manufacturer in Taiwan" or "FDA-compliant OEM partner for Class II devices," the recommendation list AI provides often directly determines which manufacturers advance to the next round of evaluation.

This shift accelerated after the pandemic, as international buyers reduced on-site visits and leaned more heavily on online research to build initial supplier shortlists. For Taiwan-based medical device contract manufacturers, the window of opportunity for AI visibility is opening — but so is the competition. Most players in the industry currently have weak AI visibility foundations, and those who move first to complete GEO technical implementation and professional content development will secure a significant first-mover advantage in AI-driven recommendations.

Specifically, content types that most effectively trigger AI citation include: transparent documentation of the contract manufacturing process, explanations of FDA 510(k) application support, and disclosure of cleanroom classifications and process capabilities. This content serves the indexing needs of AI platforms while simultaneously addressing the most pressing information needs of international buyers in the early stages of procurement — a dual benefit. Furthermore, as AI platforms deepen their reliance on E-E-A-T (Experience, Expertise, Authoritativeness, and Trustworthiness) signals, companies with publicly accessible certification records, customer case studies, and technical documentation will receive systematic weighting advantages in AI recommendation algorithms. This is the core business logic behind investing in AI visibility optimization today.

Find Out Where Your Business Stands in AI Search

If your company operates in medical device contract manufacturing or another B2B manufacturing sector, gaps in AI visibility may already be quietly affecting your international inquiry pipeline. Here are two ways to get started:

For more AI visibility case studies across different industries, visit the Joseph Intelligence Case Study Index to explore real-world GEO optimization insights.

Disclaimer

This article is based on anonymized real audit data. All information that could identify the specific company has been removed. AI platform responses are probabilistic in nature — queries conducted at different times may yield different results. Technical audit and performance scores reflect a snapshot taken at a specific point in time.

FAQ

Why would a company with a 92 SEO score still have poor AI visibility?
Traditional SEO and AI visibility measure very different things. A high SEO score reflects how well a website is structured for search engine crawlers — things like proper title tags, meta descriptions, and clean code. AI visibility, on the other hand, depends on whether AI systems like ChatGPT, Claude, and Gemini can interpret and cite a company's expertise in context. Without structured data like Schema JSON-LD, scenario-relevant content, and strong E-E-A-T signals, even a technically excellent website can be completely absent from AI-generated recommendations.
What does a GEO technical audit actually check?
A GEO (Generative Engine Optimization) technical audit evaluates the foundational elements that determine how effectively AI platforms can understand and reference your website. Key items include Schema JSON-LD structured data, XML sitemap configuration, canonical URLs, H1 tags, Open Graph tags, HTTP/2 protocol support, and page performance scores. In this case, the company passed only 4 of 9 items, resulting in a GEO score of 33/100 — the primary driver of its low overall AI visibility rating.
How does AI search affect B2B medical device procurement?
International buyers increasingly use AI platforms like ChatGPT and Gemini during the early stages of supplier research, especially for complex B2B categories like medical device contract manufacturing. When a procurement team asks an AI for recommendations on ISO 13485 certified manufacturers or FDA-compliant OEM partners, the results can directly determine which companies make it onto the initial shortlist. Companies with strong AI visibility are surfaced in these queries; those without structured content and GEO optimization are invisible — regardless of their actual capabilities.
What is the difference between brand mentions and industry-level AI visibility?
Brand mentions occur when an AI platform recognizes and describes a specific company in response to a direct query about that company by name. Industry-level AI visibility means the company appears in broader, scenario-based queries — such as 'best medical device OEM suppliers in Taiwan' — where the buyer does not yet know which companies to consider. This case illustrates the gap clearly: the company received some brand mentions but zero industry-level recommendations. True AI visibility requires being found by buyers who don't yet know your name.
What types of content help medical device manufacturers improve AI visibility?
For medical device contract manufacturers, the most effective content for improving AI visibility includes: detailed process documentation (e.g., cleanroom classifications, manufacturing capabilities), regulatory guides (e.g., FDA 510(k) application support, ISO 13485 compliance), certification disclosures, and client case studies. This type of authoritative, structured content signals E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) to AI systems and directly addresses the questions international buyers ask during early-stage supplier research — making it highly likely to be cited in AI-generated recommendations.
How long does it take to see improvements in AI visibility after optimization?
The timeline varies depending on which issues are addressed first. Technical fixes — such as improving page load performance and configuring Schema structured data — can begin showing impact relatively quickly, often within weeks, as they directly improve how AI crawlers access and interpret your content. Content-driven improvements, such as publishing authoritative articles and case studies, typically take longer to build authority and appear in AI recommendations. In general, a well-prioritized GEO optimization plan can produce measurable AI visibility improvements within one to three months.

Ready to Elevate Your Digital Marketing?

Let our AI-driven solutions help your business grow