Skip to content
CheckerGoogleRank Version française
  1. Home
  2. Technical and performance
  3. Googlebot Simulator

Googlebot Simulator

See a page exactly as a search engine crawler receives it — before rendering.

The page is requested with the chosen crawler's user agent and the raw HTML is analysed — without executing JavaScript, exactly like a crawler's first pass. This is what gets indexed immediately; rendered content arrives later, if at all. No usage limit.

What this tool is for

Your browser is generous. It runs scripts, waits for late resources, applies styles, and assembles a finished page. A crawler's first pass does none of that: it takes the HTML the server sent and reads it as text. What is not in that HTML does not exist yet.

The gap between the two is where invisible problems live. A page rich with content in the browser and nearly empty to a crawler. A security service refusing bots while every human visitor sails through. Neither shows any symptom you could notice by looking.

This simulator makes that first pass visible, for four different crawlers. The response code, the text actually extracted, the heading structure, the followable links, and the indexing directives — the raw material every engine starts from.

How to read your results

The response code comes first because nothing else matters if it is wrong. 200 is correct. 403 or 503 to a crawler while a browser gets 200 means something is filtering bots, and the page is effectively invisible to search.

The extracted text is the honest measure. If your page shows a thousand words in the browser and forty here, the content is built by JavaScript and will be indexed late — or not at all if rendering is deprioritised.

The heading structure should read like a table of contents. A single H1 naming the subject, H2 for the main sections. Several H1s or a jump from H1 straight to H4 usually means headings were chosen for their size rather than their meaning.

The followable links are how crawlers move through your site. Links created by JavaScript on click are not visible in this pass. A page whose only navigation is scripted is a dead end for discovery.

The structured data found in the raw HTML is what qualifies for rich results. Structured data injected by a script after load is far less reliably picked up.

The most common mistakes

Assuming a bot protection service is harmless. Cloudflare, Sucuri and similar tools sometimes challenge legitimate crawlers, especially after a rule change nobody reviewed. The site works flawlessly for humans while quietly disappearing from search.

Building the main content in JavaScript. It is not forbidden and Google does render — eventually, at lower priority, unpredictably. For content you want indexed quickly and reliably, server-rendered HTML remains the only dependable answer.

Hiding text behind a consent banner. Crawlers do not click "accept". If your content is only revealed afterwards, the crawler sees a banner and nothing else.

Serving a stripped-down page to bots to look faster. Substantial differences between what bots and humans receive is cloaking, and it is one of the few violations that reliably attracts a manual penalty.

Testing only Googlebot. Bingbot powers Bing and, indirectly, several AI assistants. GPTBot and ClaudeBot are how a growing share of answers get sourced. They do not always receive the same response — check the ones that matter to you.

Related tools

Frequently asked questions

Does Googlebot run JavaScript?

Yes, but in a second pass. The first pass reads the raw HTML; rendering happens later, when resources allow, and can lag by hours or days. Content that only exists after JavaScript runs is therefore discovered late and sometimes not at all. Anything that matters should be in the HTML the server sends.

Why does my page look different to the crawler?

Usually because content is injected by JavaScript, loaded on scroll, or hidden behind a consent banner that a crawler never dismisses. Occasionally because the server responds differently to a bot user agent — which, if the difference is substantial, is cloaking and carries a penalty.

Is serving different content to bots ever allowed?

Minor variations are fine: no cookie banner, no personalisation. Substantial differences in the main content are cloaking, which Google treats as a spam violation. The safe test is whether a human visitor would recognise the page from what the crawler received.

What does a 403 to Googlebot mean?

Something is refusing the crawler — usually a firewall, a bot-protection service or an over-eager security plugin. It is one of the most damaging misconfigurations possible, because the page becomes completely invisible while looking perfectly healthy in a browser. Check the rules of any WAF or CDN in front of the site.

How many words does a page need?

There is no threshold, and Google has denied that one exists. A page answering a precise question in two hundred words can outrank three thousand words of padding. What matters is whether the page answers the question someone actually asked.

Do heading levels matter?

For structure and accessibility, yes. As a ranking factor, far less than commonly claimed. One H1 per page, then H2 and H3 in a logical order — skipping from H1 to H4 confuses assistive technology more than it confuses Google.