How to use site: search to find what you already have
Before briefing another article, one operator tells you whether the site already covers the topic — and it usually does.
4 min read
Before you brief another article, the useful question is whether the site already covers the topic. A single search operator answers it in about ten seconds, and the answer is often yes. This is the site: operator pointed at your own domain with a topic attached, and it is the cheapest content audit available.
Every content plan starts with a list of things to write. Almost none of them start by checking what has already been written, which is how sites end up with three thin pages on a subject that deserved one good one.
The question this answers
Do we already have a page about this, and if so, is it the page I think it is?
Your CMS will not tell you reliably. A search of the CMS matches titles and tags, not the actual language on the page, and it has no idea which of your pages Google considers relevant to the topic. The operator asks Google that question directly, which is the opinion that matters.
The syntax
No space after the colon. The domain goes without https:// and without www unless you specifically want to exclude a subdomain. Everything after the domain is treated as ordinary query terms, so you can use quotes to force an exact phrase:
The unquoted version shows you everything Google thinks is topically related. The quoted version shows you only pages carrying that exact string. Run both — the gap between them is informative on its own.
A worked example
Take a company that sells ergonomic office furniture and is about to brief a piece on desk height. The check before writing:
Say it returns four results: a product page for a specific desk, a buying guide that mentions height in one paragraph, an old blog post titled “Finding your ideal desk height,” and a support article about the adjustment mechanism.
SCREENSHOT — Emily: the results for site:example.com desk height, showing the four pages and their titles.
That is the answer, and it changes the brief. There is already a page on this exact subject. The question is no longer what to write; it is whether the existing post is good enough to keep, and whether the buying guide should link to it instead of half-covering the topic itself.
Want to check yours?
Run the Content brief builder →
Variations worth knowing
The operator takes a path as well as a domain, which narrows the question to one part of the site:
That is useful when a topic is covered in both marketing pages and support documentation and you only want to audit one of them. It also works on subdomains, which is how you find the content nobody remembers owning:
And it combines with a file type, which finds the PDFs and spreadsheets that were published years ago and never linked from anywhere:
None of these are exotic. They are the same operator with the question narrowed, and narrowing the question is most of the skill.
What the results do not tell you
A site: search lists pages Google has indexed and considers related to your terms. It is not a complete inventory of your site, it is not ranked the way a normal search is ranked, and the order it returns things in should not be read as importance.
So treat the output as a candidate list, not a verdict. A page missing from the results might be unindexed, might use different vocabulary, or might simply not have surfaced for that phrasing. The operator is a fast way to find most of what exists, not a guarantee of finding all of it.
Reading what comes back
Three patterns are worth recognizing.
Nothing comes back. Either the topic is genuinely uncovered, or your pages use different language than you just typed. Try the phrasing a customer would use before concluding it is a gap.
One clear page comes back. The topic is covered. The work is improving that page, not adding a second one beside it.
Several partial pages come back. The most common result, and the most useful. The topic is spread across pages that each mention it without owning it.
The third case is the one that quietly costs traffic, and it is the subject of the next part in this series.
What to do with the results
Write the URLs down before you do anything else, because the search is easy to rerun and easy to forget. For each one, decide between four outcomes: keep it as the page that owns the topic, fold it into another page and redirect, leave it alone because it serves a different intent, or delete it.
Doing this before the brief rather than after publication is the whole point. Once the new page exists, you own the problem of two pages competing, and the fix costs redirects and internal link edits instead of one search.
Where this hands off
When the check shows a topic is genuinely uncovered, the next step is a brief built from the results page rather than a template — the content brief builder does that job. When it shows several partial pages, part two of this series covers deciding which one should win.
If the site is large enough that running these searches one at a time is impractical, part eight covers turning operator results into a crawl list so the same questions can be asked across hundreds of URLs at once.