Sigma-i Care AEO Assessment Report
by Novastacks AI

This is the #1 issue and the likely root cause of zero rankings. sigmaicare.com serves the default self-signed Plesk hosting-panel certificate (subject/issuer CN=Plesk, self-signed, SSL error code 18) instead of a valid CA-signed certificate. The consequences are severe and compounding:

• Real users on Chrome and Safari hit a "Your connection is not private" / "Not Secure" security interstitial before they can reach the homepage — a trust killer at the front door.
• Search engines and AI crawlers reject the HTTPS connection outright (the testing engine returns INSECURE_DOCUMENT_REQUEST; a standard secure connection fails with "self signed certificate"). A site that cannot be securely connected to is not indexed or cited.
• Core Web Vitals / performance could not be measured at all — the testing engine cannot load the page over a secure connection. The unmeasurability is itself the finding: if our tools cannot reach your site securely, neither can Google or ChatGPT.
• The one branded query where the prospect does rank organically surfaces an insecure http:// URL, not https — confirming the apex redirects into the broken certificate.

The fix is fast and high-impact: install a valid CA-signed certificate (a free Let's Encrypt certificate is sufficient) and force all traffic to HTTPS. On the positive side, the underlying site is server-rendered (Apache + CodeIgniter/PHP) and returns full HTML over plain HTTP — so once the certificate is valid, the content is immediately crawlable. This is a 1–2 week fix that unblocks everything downstream.

Across every analyzed prospect page, the on-page signals search engines and AI use to understand a page are nearly absent. The homepage carries no H1 and no H2 headings and roughly 70 words of body text; the Solutions page is similar (68 words, no H2). By contrast, tBPC's homepage uses 25 H2 sections and its product page carries 18, giving machines a clear topic map. Two further misconfigurations compound the problem: the site declares its content language as Japanese (ja) when it is Traditional Chinese, which can cause AI crawlers to misclassify and mis-serve it; and the social icons in the footer link to facebook.com / youtube.com / instagram.com root pages rather than real owned profiles.

No page on sigmaicare.com contains any structured data (zero JSON-LD blocks site-wide). Structured data is how a site tells Google and AI engines "this is an organization, this is a product, this is an article" in machine-readable form — it is also what powers knowledge panels and rich results. Its complete absence is a primary reason the prospect's own branded search shows a competitor's knowledge panel instead of its own. tBPC, by comparison, implements Organization, WebSite, Product, Article and SearchAction schema, which is part of why it is the one local competitor AI can readily describe and compare.

We tested 6 category and comparison queries as a diagnostic signal. The prospect was absent from the top 20 organic results on all three core category keywords (智慧健康手環 心律 血氧 ECG; 穿戴式裝置 健康監測 演算法 廠商; PPG 血氧 心電圖 感測器), each of which triggered a Google AI Overview. Competitor tBPC appeared on 4 of the 6 SERP tests (including #2 on the prospect's own self-description query) and Osmile on 2. The pattern: the prospect is present only when its brand name is supplied, and absent from every query a buyer who does not yet know the brand would type.

Across both ChatGPT and Google AI, every category and transactional answer was organized around one signal: government medical certification (TFDA / 衛福部 醫器 license numbers). tBPC's Soosyn ST-50 is cited repeatedly precisely because it surfaces its 衛福部 certification number; ASUS VivoWatch wins on TFDA certification. Sigma-i Care builds medical-grade sensor technology for preventive medicine — but its site surfaces no certification, no patent, and no trust pages, so AI has nothing to attach trust to. This is a positioning-and-evidence gap, not a technology gap: the substance likely exists, but it is invisible on the site.

The two competitors share a tight cluster of high-commercial-intent keywords — 'medical-certified health wristband' (醫療認證 健康 手環), 'medical-grade blood-pressure wristband' (醫療級血壓手環) — with search volumes of 90–110 each. Osmile ranks #3–#8 and tBPC #5–#36 on these. These are exactly the buying-intent terms a PPG/SpO2/ECG wristband maker should own, and Sigma-i Care ranks for none of them. Because the prospect has zero ranked keywords, there is no keyword overlap to even measure — the moat is currently 100% competitor-held.

There is a real asset to build on. On both queries where ChatGPT described Sigma-i Care, it cited sigmaicare.com directly as the source — the company's own words about self-developed PPG/SpO2/ECG algorithms and AI big-data analytics were accurate and compelling enough for the model to repeat them. The technology narrative is differentiated and the model already trusts the first-party source. The moment the certificate is fixed, schema is added, and the demand-side content exists, the prospect has a credible, AI-trusted story to rank — it is currently a distribution problem, not a substance problem.