Technical SEO for Corporate Websites: Core Web Vitals, Prerendering, and Structured Data

Modern corporate sites are mostly developed with client-side libraries like React. This provides rich interactivity; however, it requires additional technical care for search engines to view the content and for a fast page experience. This article covers the technical side, not the content and keyword strategy: is the page loading correctly, in the right language, and in the right structure?

1. Core Web Vitals: Measurable User Experience

Google's Core Web Vitals metrics consist of three measurements: LCP (Largest Contentful Paint: the time it takes for the main content to appear), INP (Interaction to Next Paint: response time to clicks), and CLS (Cumulative Layout Shift: elements jumping while the page is loading). These indicators are among ranking signals and directly reflect the user experience.

Practical improvements:

• Load hero images in modern formats (WebP/AVIF), in the correct size and with priority; let the remaining images load lazily.
• Start heavy animations and 3D scenes only if the device supports them and after the main content has loaded; respect the reduced motion preference.
• Assign dimensions to images and embedded content to prevent layout shifts.
• Split the JavaScript bundle on a route basis; the homepage should not download the admin panel code.

2. Prerendering and Server-Side Rendering

On a client-side rendered page, the initial HTML is empty; although search engines can execute JavaScript, this can be delayed and variable. The solution is to generate the HTML of public pages at build time (prerender) or use server-side rendering. For corporate sites whose content comes from a database, build-time prerendering combined with rebuilds when content changes is a practical and inexpensive solution.

// Example: prerendering the route list after build (conceptual)
const routes = ['/', '/hakkimizda', '/cozumler', ...servisSluglari.map((s) => '/cozumler/' + s)];
for (const route of routes) {
  const html = await renderRouteToHtml(route);   // including title, description and content
  await writeFile('dist' + route + '/index.html', html);
}

3. Title, Description, and Canonical Address

Every page must have a unique title and meta description tag; these fields should come from content management, not be hardcoded in the code. If the same content is accessible via multiple addresses (for example, the difference of a trailing slash), the canonical tag should point to a single address.

4. Multilingual Sites: hreflang

On a site with Turkish and English versions, each page must specify its other language counterpart with the hreflang tag; thus, the search engine presents the correct language to the user and does not consider the versions as duplicates of each other.

<link rel="alternate" hreflang="tr" href="https://ornek-site.com/tr/cozumler" />
<link rel="alternate" hreflang="en" href="https://ornek-site.com/en/solutions" />
<link rel="alternate" hreflang="x-default" href="https://ornek-site.com/tr/cozumler" />

5. Structured Data: schema.org JSON-LD

Structured data tells the search engine what the page is: organization, service, software application, article. The JSON-LD format is added independently of the page content and is easily validated. Important rule: structured data must not contain information that is not visible on the page (for example, non-existent ratings or reviews); this is both misleading and may face penalties.

{
  "@context": "https://schema.org",
  "@type": "Article",
  "headline": "Kurumsal Web Siteleri İçin Teknik SEO",
  "datePublished": "2026-07-10",
  "author": { "@type": "Organization", "name": "Cy Logic Software" }
}

6. Sitemap and Robots

sitemap.xml must only contain published pages and should be automatically generated when content changes; drafts or hidden pages should not be in the list. robots.txt excludes the management panel and API addresses from crawling; however, robots.txt is not relied upon for privacy, this is not an access control.

The goal of technical SEO is not to trick the search engine, but to present what the site actually is in a format that machines can also read.

7. Accessibility is a Friend of SEO

A meaningful heading hierarchy (single h1, ordered h2/h3), alt text on images, keyboard-navigable menus, and sufficient color contrast are valuable for both accessibility and indexing. These should not be add-on decorations, but part of the component library.

8. Monitoring

Indexing status, crawl errors, and Core Web Vitals reports should be monitored regularly with tools like Search Console. These reports should be checked after every major change made to the site (new route structure, language change).

SEO Fields Coming from Content Management

No matter how good the technical infrastructure is, each page's title, description, and canonical address must be in the hands of the content team. Practical model: next to each content record (service, product, article) there are optional SEO fields; if left blank, the system produces a reasonable default from the title and excerpt. Open Graph fields (title, description, image) for social sharing cards are managed from the same place; if there is no sharing image, the brand default is used.

Image Optimization: The Biggest Gain

The vast majority of page weight on corporate sites consists of images. Server-side sizing and modern format conversion (WebP/AVIF) at load time, device-appropriate size delivery with srcset, lazy loading of off-screen images, and giving width/height to every image directly improves both LCP and CLS. The media library should perform these conversions automatically; editors should not be expected to "upload in the correct size".

URL Structure and Redirects

Addresses must be short, readable, and permanent. When content is moved or the slug changes, the old address must be redirected to the new address with a 301; otherwise, earned rankings and backlinks are lost. The redirect table must be part of content management; every slug change should generate an automatic redirect record. On multilingual sites, the language must be part of the address (e.g., /tr/, /en/); the same address changing with cookies or browser language is problematic for indexing.

Crawl Budget and Unnecessary Pages

Filter combinations, sorting parameters, and pagination can produce hundreds of different addresses for the same content. These addresses either need to be connected to the homepage with a canonical or not indexed (noindex). The management panel, login pages, and API endpoints should not be indexed; however, their privacy should rely on access control, not robots.txt.

Measurement: Lab and Field Data

Tools like Lighthouse provide lab measurements; real user data (Chrome User Experience Report and real user measurement on your site) is the value considered in ranking. The two can produce different results: a page that looks fast in the lab may perform poorly on slow mobile connections. Adding real user measurement to your site should be evaluated in terms of personal data and kept anonymous if necessary.

// Example: real user measurement with web-vitals (conceptual)
import { onLCP, onINP, onCLS } from 'web-vitals';
const send = (metric) => navigator.sendBeacon('/api/vitals', JSON.stringify({ name: metric.name, value: metric.value, route: location.pathname }));
onLCP(send); onINP(send); onCLS(send);

Pre-launch Checklist

1. Unique title and description on every page
2. Canonical address and hreflang tags
3. Is content visible in pre-rendered HTML?
4. Did structured data pass the validation tool?
5. Does sitemap.xml only contain published pages?
6. Are images sized, in modern format, and lazy-loaded?
7. Are 301 redirects defined for old addresses?
8. Are there no horizontal scrolling and small touch targets in mobile view?

SEO-compliant architecture in Cy Logic's web and mobile applications service is not a marketing plugin, but the default delivery standard of corporate sites; our digital marketing team undertakes the content and measurement side.