One-Page Websites Are Invisible to Google and AI Search: 7 Critical Problems Found in a Real Manufacturing Audit

We recently audited a manufacturing company's website — a business with 10+ years of operations, 1,000+ completed projects, and a 700 m² production facility. Their portfolio includes public infrastructure projects in major European cities. Real expertise, real products, real clients.

Their website told a different story.

Google had indexed exactly one page. Not one category page. Not one product page. One page total — the homepage. By commercial search queries that should bring new customers ("LED lighting poles manufacturer," "park lighting poles price," "street lighting production company"), the website was nowhere in the top 10. Competitors with objectively weaker portfolios occupied every result on page one.

And in AI search — ChatGPT, Perplexity, Google AI Overview — the company simply did not exist.

This article breaks down the 7 critical problems we found, explains why each one matters for SEO, AEO, and GEO in 2026, and maps each problem to the specific service that fixes it. No client names — just the anatomy of a website that's leaving money on the table.

In 2026, there are three search channels, not one:

SEO (Search Engine Optimization) — classic Google and Bing rankings. Still the largest traffic source, but declining as a share of total discovery.

AEO (Answer Engine Optimization) — optimization for Featured Snippets, Google's "People Also Ask" boxes, and voice assistants. When someone asks "what is the best type of lighting pole for a park," Google pulls the answer from a website — but only if it's structured correctly.

GEO (Generative Engine Optimization) — optimization for AI-generated answers in ChatGPT (900M+ weekly active users), Perplexity, Google AI Overview, Gemini, and Claude. These systems cite websites that provide clear, structured, factual content. If your website isn't optimized for extraction, AI will cite your competitors instead.

The website we audited was failing on all three.

The entire website was a single-page landing with anchor navigation. The menu items ("About Us," "Projects," "Technical Specs," "Portfolio") were not separate pages — they were #hash links scrolling to sections of the same URL.

Google indexes pages, not sections. When the entire website is one URL, Google can only rank that URL for one primary keyword cluster. But a manufacturing company needs to rank for dozens of distinct queries: conical lighting poles, telescopic poles, park lighting, highway masts, LED streetlights, custom metalwork. Each needs its own page with its own title, H1, meta description, and Schema Markup.

We confirmed with a site: operator search — Google's index contained exactly one result. Competitors in the top 10 all had multi-page websites with separate URLs per product category, each containing 800–2,000 words of optimized content.

The impact: The website could only rank for its own brand name. Every non-branded commercial query went to competitors.

Expanding from 1 page to 20+ without a content strategy leads to random pages targeting random keywords. Before development, build the keyword and URL foundation:

The main heading was split into two separate <h1> tags for a visual animation effect. Additionally, the site used <h6> tags for descriptive subtitles that should have been <p> or <h2>, creating a broken hierarchy (H1 → H6, skipping H2–H5).

The H1 tag is the single most important on-page SEO signal — it tells Google what the page is about. When H1 says "Why" and another H1 says "us?", Google sees two competing topics, neither of which is a searchable keyword.

For AEO, heading structure is how Google extracts content for Featured Snippets. A clear hierarchy (H1 → H2 → H3) with keyword-rich headings is the foundation of answer extraction. For GEO, AI crawlers use headings to understand content structure. Broken hierarchy = broken extraction.

The correct H1 should target the primary commercial query with location and service: "LED Lighting Pole Manufacturer — Production, Design, Delivery." One H1, one clear topic, one keyword cluster. For local markets, add the city or region.

Over 30 images — company logo, project photos, portfolio gallery — all with empty or missing alt attributes. File names were generic: 1.jpg, 2.jpg, gallery/11.jpg. Format was JPG/PNG with no WebP conversion, no lazy loading, no srcset for responsive delivery.

For SEO: Google Images is a significant traffic source for manufacturers — architects, designers, and engineers search by image. Without alt text and descriptive filenames, these images are invisible to Google's image index.

For GEO: An image with alt="Conical lighting pole 8m installed in a city park — custom production 2024" gives AI a citable fact. An image with alt="" gives it nothing.

For performance: Unoptimized images increase mobile load time by 2–5 seconds, directly hurting Core Web Vitals — a confirmed Google ranking factor since 2021.

The website had zero structured data. No Schema.org markup of any kind — not Organization, LocalBusiness, Product, FAQ, BreadcrumbList, or Review.

Schema Markup is the language that tells search engines *what* your content means, not just what it says. Without it:

- Google doesn't know this is a manufacturing company — can't show the Knowledge Panel with address, phone, hours, and logo - No FAQ rich results (FAQPage schema) — no expandable Q&A in search results, which get 2–3× higher click-through rates - No review stars (AggregateRating) — even though the company claims 1,000+ satisfied clients, there's no structured proof - No breadcrumbs (BreadcrumbList) — search result shows a raw URL instead of a navigational path

For GEO: Schema Markup is one of the primary data sources AI uses to build entity graphs. When ChatGPT is asked "who manufactures LED lighting poles in Europe," it looks for Organization schema with makesOffer, areaServed, address, and foundingDate. Without Schema, the company doesn't exist in the AI's entity graph.

The website had no llms.txt or llms-full.txt files, no AI bot access configured in robots.txt, no Quick Answer Blocks, and no content structured for AI extraction.

When we tested queries like "best LED lighting pole manufacturers" in ChatGPT and Perplexity, the company was not cited in any response. Competitors with proper AI optimization were.

The paradox: This company had *exactly* what AI loves to cite — real technical specifications (IP67, Osram/Cree/Seoul LEDs, 50,000-hour lifespan), real project locations (named cities and landmarks), concrete numbers (4,760 poles manufactured, 10-year warranty). But none of it was formatted for extraction.

Quick Answer Blocks are 40–60 word factual paragraphs following the pattern: "What is [topic]? [Topic] is [concise definition]. Key features include [fact 1], [fact 2], [fact 3]." AI systems parse and cite this format readily. Without it, the expertise exists but remains invisible.

No publicly accessible robots.txt or sitemap.xml were found. The website was not connected to Google Search Console or Bing Webmaster Tools.

Without sitemap.xml, Google has to discover pages by crawling links — unreliable for a single-page site with no internal links to other pages. Without robots.txt, search engines and AI bots have no instruction set.

Without Google Search Console, the website owner had zero visibility into which queries bring impressions and clicks, which pages are indexed, crawl errors, mobile usability issues, Core Web Vitals scores, or manual actions. This is the equivalent of running a business without looking at the bank account.

No Open Graph tags (og:title, og:description, og:image) and no Twitter Card tags were present. The meta author tag was empty. No canonical URL was specified.

When someone shares the website link in Facebook, LinkedIn, Telegram, Viber, or WhatsApp, the preview shows nothing — no image, no description, no reason to click. In B2B manufacturing, where deals often start with a link shared in a work chat, this is a missed conversion opportunity on every share.

The missing canonical tag is also a technical SEO issue — without it, search engines may index duplicate versions of the same page (with/without trailing slashes, with/without www), splitting ranking signals.

Here's how we sequence the work, organized by impact and dependency:

Based on our experience with similar B2B manufacturing websites:

You can run a basic self-audit right now: