Deutsch
How search engines render

Web Rendering Service

Also known as: WRS

The Web Rendering Service (WRS) is the component of Google Search that renders pages and executes their JavaScript. It works after crawling and supplies the rendered HTML that Google uses for indexing. Google describes the renderer as a headless Chromium and says that Google Search runs JavaScript with an evergreen version of it.

How does rendering work at Google?

Googlebot first fetches the URL. Pages with a 200 status code are queued for rendering unless a robots meta tag or header forbids indexing, and the wait can last a few seconds or longer. When Google’s resources allow, a headless Chromium renders the page and runs the JavaScript. Google does not render pages or files blocked by robots.txt.

Google’s troubleshooting guide lists behaviors of the WRS to design for:

  • It loads each URL like a regular browser and follows server and client redirects, but local storage, session storage and cookies are cleared across page loads.
  • It may ignore caching headers and use outdated JavaScript or CSS files. Content fingerprinting in file names, like main.2bb85551.js, avoids this.
  • It does not support WebSockets or WebRTC connections, so content needs an HTTP fallback.
  • It flattens the light DOM and shadow DOM.
  • Googlebot and the WRS may skip resources that do not contribute to essential page content, so client-side analytics may not show their activity accurately.
From URL to search resultFive stations from left to right: discovery via links and sitemaps, crawling after checking robots.txt, rendering after waiting in the render queue, indexing with duplicate clustering and canonical selection, serving matching results. Links found in the rendered HTML lead back to discovery. Below each station are typical hurdles according to Google.1 · CRAWLING2 · INDEXING3 · SERVINGDiscoverlinks onknown pages,sitemapsCrawlcheck robots.txt,then fetchthe URLRenderrender queueChromium runsthe JavaScriptIndexcluster pages,pick canonical,storeServeresults fromthe index forthe querynew links from the rendered HTMLTYPICAL HURDLES ACCORDING TO GOOGLEno link,no sitemap*server ornetwork errors,robots.txt blocksno 200 statusor noindex: mayskip renderinglow quality,noindex, duplicateof another URLnot relevant,low quality,robots meta ruleSchematic. Google counts discovery and rendering as part of the crawling stage.* An inference, not a Google statement.
Schematic path of a URL through Google Search. Google describes three stages: crawling (including discovery and rendering), indexing and serving search results. Indexing is not guaranteed. Sources: Google Search Central: In-depth guide to how Google Search works; Google Search Central: Understand JavaScript SEO basics.

Example

A shop loads product texts from /api/, and robots.txt blocks that path. The texts are missing from the rendered HTML. The URL Inspection tool in Search Console shows the rendered result and reveals the gap.

More on the searchVIU blog

Sources

  1. Fix Search-related JavaScript problems – Google Search Centraldevelopers.google.com
  2. Understand JavaScript SEO basics – Google Search Centraldevelopers.google.com

Last updated: