View a Page as Googlebot
Request the same page twice, once as a normal phone visitor and once announcing itself as Googlebot, and compare what each one is served.
What the two requests compare
Both requests use the same phone and browser string. The only difference is the Googlebot token, so any change in the answer comes from the site reacting to that name.
- robots.txt for Googlebot
- The rule that applies to Googlebot for this exact address, and whether it came from a Googlebot group or the group for all crawlers. If Googlebot is disallowed, its request is not sent at all.
- Status and final address
- The HTTP status and where each request ended up after redirects.
- Title, description and canonical
- The three head tags most often changed by cloaking, geo redirects or misconfigured caching.
- Meta robots and X-Robots-Tag
- Indexing directives from the page head and from the HTTP header, including any written only for googlebot.
- Headings
- The first H1, the number of H1 tags and the number of H2 tags.
- Structured data types
- The schema.org types found in the JSON-LD of each answer.
- Words and links
- Word count, internal links and external links. Small gaps from dates or tokens are ignored, so only real changes are listed.
- Text before JavaScript
- The visible text the HTML delivers on its own, which shows how much of the page exists before any script runs.
How to see a page as Googlebot
Paste the page address
Use the exact URL you want to rank, including any path and query string, since robots.txt rules are matched against both.
Run the check
robots.txt and the phone browser request go first, then the Googlebot request if robots.txt allows it.
Read the differences table
Each row shows a signal with the browser value on one side and the Googlebot value on the other.
Confirm in Search Console
For anything that looks wrong, run URL Inspection on the same address to see the page Google itself fetched and rendered.
Each request waits about 10 seconds and reads up to 2 MB of HTML, following at most 5 redirects.
Reading the side by side result
- As A Phone Browser / As Googlebot's Name
- Whether each request was served, with its status, or the reason it failed.
- Signals That Differ
- How many compared signals changed. Zero means the site does not change this page by User-Agent.
- robots.txt For Googlebot
- Allowed or not, with the deciding Allow or Disallow line. "Not requested" means Google would not crawl this address.
- Googlebot refused with 401, 403, 406, 429 or 503
- Usually a firewall rejecting a Googlebot name from a non-Google address. That blocks impostors, not the real crawler.
- Very little text in the HTML
- The page probably builds its content with JavaScript. Google renders scripts, this tool does not, so check the rendered HTML in URL Inspection.
- Notes
- Plain statements of what the result can and cannot prove for this page.
What a Googlebot view often uncovers
A different title or canonical served to Googlebot
Fix: Look for a plugin, CDN rule or server config that branches on User-Agent. Serve crawlers and visitors the same head tags unless you have a documented reason.
A noindex that only appears for Googlebot
Fix: Some security and staging plugins add robots directives for bots. Remove the rule, then request indexing again in Search Console.
robots.txt disallows a page you want ranked
Fix: Edit the Disallow line the tool quotes, or add a more specific Allow for that path. Test the change before publishing.
The HTML holds almost no text
Fix: Render the main content on the server or pre-render it, so the words exist in the first response and do not depend on script execution.
Fewer internal links for Googlebot
Fix: Menus or related links injected only for visitors, or hidden behind bot detection, starve deeper pages of crawl paths. Put key links in the served HTML for everyone.
Who uses a Googlebot viewer
- SEO consultants investigating a page that dropped out of the index for no clear reason.
- Developers testing that a bot protection rule does not change content for crawlers.
- Site owners who suspect a hacked site is showing spam only to search engines.
- Teams moving to a JavaScript framework who need to know what exists before rendering.
A User-Agent test, not an address test
Google recognises its crawler by the IP address a request comes from, confirmed by reverse DNS, not by the name in the header. A site that cloaks by address will serve this tool the visitor page. So a clean match proves the site does not branch on the name, and nothing more. Hacked sites in particular often check the address, which is why Search Console URL Inspection stays the final word.
Bot Viewer FAQ
Short answers about the Bot Viewer.
No. It sends Googlebot Smartphone's User-Agent from our server, not from Google's addresses. It shows how the site reacts to the name. URL Inspection in Search Console shows the real fetch.
Most firewalls refuse requests that claim to be Googlebot but do not come from Google's network. That protects the site from fake bots. It does not mean the real Googlebot is blocked.
Neither request runs scripts. You see the raw HTML each one received. Google does render JavaScript later, so a script-built page may index more than is shown here.
robots.txt tells Googlebot not to fetch that address. Google would not crawl it either, so the tool shows the blocking rule instead.
It means the site does not change the page based on the User-Agent. Cloaking that checks the visitor's IP address cannot be seen by any tool outside Google's network.
Googlebot Smartphone, the crawler Google uses for mobile-first indexing. The browser request uses the same phone string without the Googlebot token.
Related SEO tools
Robots.txt Analyzer
28 checks on a live robots.txt file.
1 free run a dayOn-Page Audit Pro
70 checks on one page, scored against the basics.
1 free run a dayBulk Meta Extractor
Titles, descriptions and tags of 20 pages at once.
1 free run a dayAI Crawler Access Check
Which AI and search crawlers can reach your pages.
1 free run a day