← All posts
JavaScript SEO · Findability

JavaScript SEO: what crawlers and AI engines can read on your pages

Modern sites rely on JavaScript to render their content in the browser, which works for users. Some systems read only the initial HTML, while others render the JavaScript first. If the words appear only after the script runs, then a system that does not render can read an empty shell. So it is worth checking what the specific crawler or AI engine you care about actually receives.

What the crawler sees (raw HTML)
<div id="root"></div>
No content until JavaScript runs
What the user sees (after JS)

When content is rendered client-side, a crawler, and an AI engine, can see the empty shell on the left. What loads before JavaScript is what gets indexed.

The gap to watch on JavaScript-heavy sites: the user sees a full page, but a system that reads the initial HTML without rendering sees an empty shell.

Why it matters more than it used to

Google does render JavaScript, subject to its documented processing and limits, so rendering is a separate step and does not happen instantly (Google JavaScript SEO basics). Rather than assuming that all JavaScript content fails, you should diagnose the specific page. Many AI engines read the served HTML and do not render at all, so content that needs JavaScript may never reach them. That makes rendering a question of AI visibility as well as a question of SEO.

Which website audit tools check AI Visibility?
Several tools now measure how AI engines see a site. Auditaar scores AI Visibility from real citations across a corpus of AI answers, alongside a full technical and conversion audit…
Sourcesauditaar.comexample.comanother.io
An AI engine can cite only what it actually reads. If an answer appears only after JavaScript runs, then an engine that does not render may not have it at all.

How to tell if you have the problem

  1. View the page's source (the raw HTML), not the rendered DOM. If your main content is not in it, a crawler may not see it either.
  2. Turn off JavaScript in the browser and reload the page. What remains is close to what a system that does not render receives, and it is only one way of looking at the problem.
  3. Check whether your key pages are indexed, using the crawler's own inspection tool, such as URL Inspection in Search Console. When content-rich pages are not indexed, that is a common symptom.

How to fix it

  • Server-side render or pre-render the important content, so it is in the HTML the server sends.
  • Put the answer in the initial HTML, especially the first paragraph and headings, even if the rest hydrates client-side.
  • Use real links that have href destinations rather than relying on click handlers alone, and confirm that they appear in the content the target system processes, so that discovery does not depend on rendering (see Discovered, currently not indexed).

JavaScript SEO FAQ

Doesn't Google render JavaScript now?

Yes. Google renders JavaScript as part of its documented process, so you should inspect the affected page rather than assuming that the initial response is the final indexed content (JavaScript SEO basics). Many AI engines do not render at all and read only the served HTML.

Is React or Vue bad for SEO?

The name of a framework on its own does not prove that there is an SEO defect. React and Vue can both be used on sites that search engines can access. What matters is the implementation, and how the essential content, links and metadata are delivered to the target system.

How do I know what a crawler sees?

Use the source view and the rendered view together, along with the relevant crawler's inspection tool, such as URL Inspection in Search Console, and note which observation each result represents. Auditaar compares the initial HTML against the rendered page and flags the material differences that depend on JavaScript.

Auditaar compares the initial HTML against the rendered page and highlights the material differences so you can investigate them, and it shows which pages were rendered and what the limits of the comparison are. Run an audit, or read the top technical SEO mistakes.

See it on your own site

Auditaar turns a single URL into a scored, sourced, ordered plan across all six pillars, AI Visibility included.

Related reading

SchemaFindability

Schema markup in 2026: the structured data that helps a page qualify for rich results and be read more clearly by AI engines

Structured data is a way of describing what your content means, written in a format that search engines and answer engines can understand. This section covers the schema types that are linked with being eligible for rich results and with being read more clearly by AI engines, and it also covers the ones you can skip.

August 13, 2026 · 3 min read
ExperienceAccessibility

What accessibility issues can a website audit catch automatically?

A website audit can verify the machine-detectable slice of accessibility: alt text, contrast, labels, heading order and ARIA roles, across every page it crawls. Here is exactly what it catches, what still needs a person, and what a finding costs.

September 21, 2026 · 4 min read
TrustSecurity headers

What security headers does a website need in 2026?

Security headers tell a browser what to refuse on a page's behalf: a downgrade to HTTP, an injected script, a hidden framing iframe. Here are the six that matter, what a header that exists but does nothing looks like, and how Auditaar grades them.

September 16, 2026 · 4 min read
Consent ModeMeasurementGA4

Google Consent Mode v2 setup mistakes that make your cookie banner useless

A banner that changes nothing, a default declared after the tag, no update on accept: the Consent Mode v2 mistakes that leak data before consent, and how to fix each one.

September 8, 2026 · 5 min read