- There are over 130 trillon (10¹²) documents on the web (as of Jully 2016).
- Googlebot no longer crawls hashbang URLs.
- It is recommended to detect Googlebot on the server by user-agent string and send a complete “dynamically rendered” HTML document back.
- There are tools available for dynamic rendering, such as Puppeteer or Rendertron.
- If your website uses lazy loading images, add a
noscripttag around a normal image tag to make sure Googlebot will pick them up.
- Googlebot does not index images that are referenced through CSS.
- Googlebot crawls and renders pages in a stateless way, it does not support Service workers, local and session storage, Cookies, Cache API and etc.
- Google plans to bring rendering closer to crwaling and indexing and make Googlebot use a modern version of Chrome.