Google Site Search

Think of Google's index as a library card catalogue. The site: operator is like asking the librarian, "Show me every book from this author." It only returns books the library has catalogued — not every book the author ever wrote. If a page isn't indexed, it won't show up, no matter how many times you search. That is the single most important thing to understand: site: is a window into the index, not the site itself.

You would not diagnose a site's health based on site: alone, but it is the fastest check you can run without a crawl tool. I use it daily to spot whether a new page has been picked up, or to see if a competitor is ranking for a term I thought they had dropped. It is also the cheapest competitive research available — zero cost, zero logins.

The syntax is straightforward: site:example.com returns all indexed pages from that domain. Add a keyword: site:example.com SEO guide narrows results to pages Google thinks are about SEO guides. Use quotes for exact phrases: site:example.com "content audit" reduces noise. Combine with intitle: to find pages with a specific word in the title: intitle:contact site:example.com. Or filetype:pdf site:example.com to locate PDFs. These combos turn a blunt tool into a precise one.

But know the limits. The operator does not show page titles, meta descriptions, or internal linking structure. It does not reveal why a page ranks where it does. And it only reflects what Google has chosen to index — pages blocked by [robots.txt](/robots-txt/), noindexed, or behind login walls will be invisible. For a complete picture, pair site: with your [sitemap](/sitemap/) to compare what you intended to be indexed against what actually is.

Content audits: I recently ran site:myclient.com and saw 340 results. Their sitemap had 410 URLs. That 70-page gap meant 70 pages were either not indexed or not discovered. I used site:myclient.com intitle:blog to find which blog posts had made it — only half of them. That told me where to prioritise crawling or internal linking fixes. A proper [SEO audit](/seo-audit/) would dig deeper, but site: gave me the heads-up in 30 seconds.

Competitive research: For a client in travel insurance, I checked site:competitor.com and saw they had indexed 1,200 pages. But when I added site:competitor.com "travel insurance", only 40 showed. That told me they had not optimised most of their pages for that key term. I used site:competitor.com filetype:pdf to find downloadable guides — a content format they were using that my client was not.

Debugging visibility: After a site migration, site:newdomain.com is the first thing I check. If the old domain still shows 500 pages and the new one shows 50, the 301s are not working, or Google has not recrawled. I once saw a client lose 80% of indexed pages post-migration because they forgot to update internal links — site: caught it in 30 seconds.

Finding specific assets: Looking for a contact page? site:example.com contact | about. Need a list of all PDF case studies? site:example.com "case study" filetype:pdf. This is faster than any site search bar.

Edge case: The operator works on subdomains too. site:blog.example.com shows the blog index separately. Useful for segmenting large sites. But it does not work on URL fragments. And if a site uses subdirectories for different languages, site:example.com/en/ narrows to that folder.

Also, the operator can return results from different Google country versions. If you suspect geo-filtering, try site:example.com on google.co.uk versus google.de. For sites with a lot of thin pages, watch out for [duplicate content](/duplicate-content/) issues — site: will surface them if they're indexed, but not always clearly.

I used to treat site: as a complete inventory. When a client asked "how many pages do we have live?" I would run site: and quote that number. Then I found a set of pages that had been live for months but never indexed — they were hidden behind a [JavaScript login wall](/javascript-seo/). Now I always cross-reference with a sitemap and a crawl.

I also searched too broadly. Just site:example.com returns everything, including garbage pages like tag archives or thin affiliate pages. I now start with site:example.com intitle:keyword to cut noise.

Syntax is finicky. Extra spaces break the operator. site: example.com (space after colon) returns nothing useful. And you cannot combine site: with other operators inside a path — site:example.com inurl:blog works, but not site:example.com/blog.

I thought site: showed all URLs that Google knew about, including those from backlinks or sitemaps. But it only shows indexed pages. I once reported a client had 500 pages indexed, but their crawl showed 600. The missing 100 were pages Google had decided were low-quality and dropped. The client fixed the content, resubmitted via sitemap, and within weeks site: showed 580.

Also, I used to forget that site: results are not ordered by importance — just roughly by relevance to the overall query. I wasted time scrolling through page 10 of results looking for a specific URL. Now I combine with inurl: to jump straight to it. And I always check [canonical tags](/canonical-tag/) separately: if a paginated series has canonicals pointing to page 1, site: may only show that page — not the full set.