Optima Optoelectronics AEO Assessment Report
by Novastacks AI

Optima's site is server-rendered HTML on Apache with Bootstrap and jQuery. Bots can read every page. This is a real advantage in this category — Winstar runs aggressive Cloudflare bot protection that returns a challenge page to every non-browser request, including Googlebot, Bingbot, and Screaming Frog. AI crawlers face the same wall. Despite that, Winstar still earns more AI citations because their organic SEO and Knowledge Panel dominance pre-existed the bot wall. Optima's Lighthouse Performance score is 90/100, LCP is 1.3s, CLS is 0 — measurably the cleanest technical foundation of the three. The lesson: clean crawlability is a foundation, not an outcome. The site is readable; it just has nothing schema-tagged for machines to extract structurally.

Across all five Optima pages crawled (homepage, about, custom LCD module, TFT display, OLED display), there are zero JSON-LD blocks. No Organization, no WebSite, no BreadcrumbList, no Product, no FAQPage. Vitek has BreadcrumbList + ListItem schema on its product pages — the only structured-data signal in this competitor set. This is why ChatGPT's branded answer for Optima reads as a generic paraphrase: there is no machine-readable identity block for the AI to pull. Adding Organization + WebSite + Product + BreadcrumbList schema is a 1-2 week build that unlocks knowledge-graph eligibility and rich-result surfaces immediately.

Optima's homepage uses three H2 headings (What's New, Products, Optima — your optimal partner). But the TFT Display, OLED Display, Custom LCD Module, and About Us product pages each have one H1 and zero or one H2 headings. AI crawlers parse pages by heading hierarchy; without H2s, every product page reads as a single block of unstructured text. Vitek shows the same problem (most product pages have just an H1) — but Vitek at least has BreadcrumbList schema to compensate. Adding H2-level subsections ('Available Sizes,' 'Custom Options,' 'Industrial Applications,' 'Spec Sheet') turns each page into a set of extractable content blocks. Same content, restructured for machines. Also note: the mobile subdomain m.optima-opto.com.tw is where most ranking happens — search engines may be treating the desktop and mobile subdomains as two separate sites, splitting authority. Consolidating to a responsive single-domain or canonical pointing is a high-leverage technical fix.

We tested 6 category, comparison, and long-tail queries on Taiwan Google. Optima ranks #2 organic for 'Taiwan TFT LCD module supplier recommendation' — strong. Optima ranks #6 for the 'TFT LCD HDMI control board custom' long-tail with the exact-match product URL — they own this intent on their own page. But on the highest-AEO-opportunity query (OLED + E-Paper Taiwan manufacturer, which triggers AI Overview), Optima is absent from the top 12 despite selling both product lines. Winstar ranks #2 — with a dedicated OLED-specific landing page (winstar.com.tw/zh-tw/products/oled-module.html). The pattern: Optima ranks where they happen to have a product page named after the search query. Winstar ranks where they built a topical landing page around the buyer's question. The fix is not 'more pages' — it is 'pages structured around buyer questions.'

Across Taiwan Google, Winstar ranks for 135 keywords — Optima ranks for 16, an 8.4x gap. Estimated traffic value is 63x larger ($3,660 vs $58). Winstar holds 7 keywords at position #1, 11 at positions #2 to #3, and 34 in the top 10. Optima holds zero at #1, zero at #2 to #3, and one in the top 10. This is not a brand-awareness problem — Optima's brand SERP is healthy. It is a topical coverage problem: Winstar built a product taxonomy that ranks across LCM, OLED, TFT, e-paper, smart display, custom design, and accessory categories. Optima covers a similar product range in the catalog but the pages are not optimized to rank for the specific Taiwan-buyer queries. Vitek, despite being a real competitor in name recognition, ranks for only 4 keywords — they are even less topically present than Optima.

Optima's top-ranked Taiwan keywords are 'oled' (volume 6,600), '電子紙' (E-Paper, volume 5,400), '電子紙顯示器' (E-Paper display, 1,300), and 'tft lcd' (880). These are generic high-volume technology terms — Optima ranks at positions 23 to 75, meaning the pages are indexed and have authority signal but aren't optimized to compete for position #1 to #10. Winstar wins position #1 on most product-specific buyer terms. Optima's win condition: pivot the existing OLED, TFT, and E-Paper product pages into topical landing pages that answer the buyer's question ('what is a TFT LCD module,' 'OLED vs E-Paper for industrial display,' 'how to spec custom LCD'). The pages already rank for the terms — they just need to convert from product-spec pages into question-answering pages to climb past position #10.

Optima publishes in English, Traditional Chinese, and Spanish — hreflang is configured on 4 of 5 pages (missing only on the desktop homepage). Vitek publishes in 5 languages (English, Chinese, Russian, Japanese, Spanish). Winstar publishes in 2 (English, Chinese). In our AI tests we ran Mandarin queries only — but Optima's Spanish footprint is a latent asset for Mexico and South America industrial buyers, a market Winstar does not target. To convert language coverage into AI visibility, the Spanish pages need their own structured data and at least one Spanish-language piece of educational content ('Guía de selección de módulos LCD personalizados'). The infrastructure is in place; only content + schema are missing.