Skip to content
search-engine-optimizedby Netsolution

JavaScript and engine-side rendering

A JavaScript page goes through three phases: it is crawled, then queued to be rendered by a headless browser, then indexed from the result. Rendering is deferred and its delay is not guaranteed, which is enough to explain most missing content.

What exactly happens to a JavaScript page?

The guide on how a search engine works describes three stages: crawl, index, serve. For a page whose content depends on scripts, an extra phase slots in, and it is documented by name.

Processing a JavaScript page happens in three phases: crawling, rendering by a headless browser that runs the script, then indexing of the HTML obtained.

officialSource : Google Search Central, “Understand the JavaScript SEO basics”

What counts here is not the existence of rendering — it exists, the scripts are run — but the fact that it is a separate phase. Crawling downloads a file; rendering produces the real document; indexing works on the second. Three moments, three chances to lose something.

A photographic darkroom from 1956, empty: work surfaces, developing trays, an enlarger and boxes of paper stacked on shelves.
The film leaves the camera already carrying the image, but nobody can read it before it passes through this room. A page whose content comes from a script is in the same state: written, and not yet visible.USDA Forest Service, Pacific Northwest Region — Domaine public · Wikimedia Commons, relevé le

Why is rendering deferred, and by how much?

A page may stay in the render queue for a few seconds, but it can take longer than that.

officialSource : Google Search Central, “Understand the JavaScript SEO basics”

That is the most useful sentence in the whole documentation on the subject, and the least quoted. It gives an order of magnitude, then takes it straight back: “a few seconds, but it can take longer”. No delay is guaranteed, and the reason is explicit:

Rendering happens when the engine's resources allow it: crawling is immediate, running the script is not.

officialSource : Google Search Central, “Understand the JavaScript SEO basics”

Rendering is an expensive operation: it demands a full browser per page. So it is natural that it should be scheduled rather than immediate. The consequence to keep is that content produced by a script enters the index with a lag whose length is not yours to set.

Content that only exists after a script runs reaches the index later than the same content served in the initial HTML, and that gap is the real cost of deferred rendering.

hypothesis

Where does script-produced content actually get lost?

Four places, from the most frequent to the most subtle.

Blocked resources. A script essential to the content, but disallowed, will not be downloaded — and therefore not run. The page is rendered without it.

A page blocked in the robots.txt file is not downloaded, so no script it contains will be run.

officialSource : Google Search Central, “Introduction to robots.txt”

Instructions read before rendering. A no-index instruction present in the initial HTML acts before the script has any chance of removing it.

Pages that respond correctly are put in the render queue, unless a no-index instruction keeps them out of it.

officialSource : Google Search Central, “Understand the JavaScript SEO basics”

This is a classic trap for applications that render an empty shell, put a cautious instruction in it, then replace it once the data has loaded. The sequence works for a visitor and fails for a crawler, because the two do not read at the same moment.

Content that depends on an action. Anything appearing only on click, hover or scroll does not exist for an automated render, which does not interact with the page.

Addresses that do not exist server-side. Internal navigation that changes the display without changing the address produces no page to index. Symmetrically, an address that exists only as a fragment is not an address to an engine.

Serving the essential content in the initial HTML

hypothesis

Effort
high
Time to effect
weeks
Risk
none
Durability
structural

On the eight pages of this site measured on 2026-08-16, the title and the body of the article are present in the document served, with no script run at all: the only script tags carry structured data markup.

testedProtocol : Download the page with a client that runs nothing (curl), then count the script tags and check that the h1 and the body of the article are present in the document received. Eight pages measured on 2026-08-16; procedure and raw results in section C.1 of the public register.

How do you check, without telling yourself stories?

Your browser is the worst diagnostic tool available: it runs everything, immediately, with your connection and your cache. So it systematically shows the best case.

The useful comparison is between two documents. The first is the HTML served by the server, before anything runs — that is what crawling retrieves. The second is the HTML after the scripts run, as the Search Console inspection tool reports it: that is what indexing works on. If the essential content only appears in the second, the page depends on rendering, with the delay that implies. If it appears in neither, the problem is elsewhere and it is more serious.

What does not work, or no longer works

Google has run JavaScript perfectly for years: server-side rendering is no longer needed, and a single-page application ranks exactly like a classic site.

The first half is true and the conclusion does not follow. Execution exists, but it is deferred with no guaranteed delay and depends on resources nobody controls at will. Content that only exists after rendering is not in the same situation as content served from the start.

The reasonable position is neither “JavaScript breaks search” nor “the subject is settled”. It consists of deciding, page by page, what has to be readable without running anything: the direct answer, the headings, the navigation links, the structured data. The rest can depend on a script without harm. That split is an editorial choice disguised as a technical one, and that is why it is so often left to the tooling.

Sources

  1. Understand the JavaScript SEO basics — Google Search Centralaccessed on
  2. Introduction to robots.txt — Google Search Centralaccessed on
  3. In-depth guide to how Google Search works — Google Search Centralaccessed on
  4. Block Search indexing with noindex — Google Search Centralaccessed on

Block revalidated on

Frequently asked questions

Does JavaScript prevent indexing?
No, the documentation describes rendering that runs the scripts. What it introduces is an extra delay and an extra dependency: the content only exists once the script has run, and that run happens when the engine's resources allow, not immediately.
Should you move to server-side rendering?
It helps when a page's essential content depends on a script, and it is superfluous when that content is already in the initial HTML. The question to settle is editorial rather than technical: what, on this page, has to be readable without running anything?
How do you check what the engine actually sees?
By comparing the HTML served by the server with the HTML obtained after the scripts run, then inspecting the address in Search Console, which shows the render kept. A correct display in your own browser proves nothing: it runs everything, with no delay.
Why does my content appear then disappear from the results?
A gap between crawling and rendering produces that symptom: the initial HTML is indexed first, the rendered result afterwards. If the two differ a lot, the index reflects now one and now the other, until the second replaces the first.