How do I scan a site that builds its pages with JavaScript?

When you need it

Sites built with React, Vue, Angular, Next.js, Nuxt and some shop or site-builder themes send a page whose text is added by scripts. A plain scan then sees pages with little text and few links. If your pages look full in a browser but thin in the scan, switch JavaScript rendering on.

Switch it on

Website health: next to Scan again, tick Render JavaScript on (owners, admins and SEO managers). It applies from the next scan.

• Beta with Ask for early access: SEOmise has not switched the feature on yet.
• A note that it cannot run on this server: SEOmise's server cannot start the browser right now; scans read pages as their server sends them.

What a scan does then

• It reads every page as before.
• Its first pages (closest to the home page first, 50 unless SEOmise changes it) are opened in a headless browser, a little later on the queue, and read again: title, description, headings, text, links and language versions come from the rendered page.
• Links that only appear after the scripts ran are followed like any other link.
• The scan finishes when those pages are rendered (or after a time limit), and says how many pages were rendered and how many get most of their text from JavaScript.
• A page that cannot be rendered in time keeps what the plain read found.

Safe by design

Every request a rendered page makes is checked before it is made: only public addresses (never SEOmise's own network or yours), no downloads, and limits on time, memory and data per page. Pictures, video and fonts are left out because they do not change the text.