How to Do an SEO Content Audit
On this page
A content audit ends in one decision per URL: keep, improve, consolidate, redirect, noindex or remove. You reach it by joining performance, link and intent data for every URL and assigning exactly one action. The order of the questions matters as much as the data. Ask only “does it get traffic?” and you risk deleting pages other sites link to. Ask only “does it have links?” and you risk removing pages that bring in search traffic but hold no links.
Join the data before you decide anything
No single metric decides a URL. Before assigning any verdict, build one inventory row per indexable URL with at least these fields:
- clicks, impressions and average position from Search Console;
- referring domains pointing at the URL;
- word count and last-updated date;
- crawl and index status (indexed, or excluded with the reason);
- the query intent the page targets, and whether you still serve it.
Links belong in the join because they change the verdict. A page with no search traffic looks like an obvious removal until you see that other sites link to it. Google’s redirects documentation says a permanent redirect is a signal that the target should be canonical, and its guide to consolidating duplicate URLs describes canonicalization as helping search engines consolidate signals, such as links to individual URLs, into a single preferred URL. A page that returns 404 gives those links nowhere to go.
The verdict comes from three questions together: does the page earn traffic, does it hold links, and does it still serve an intent you want to own? A page can fail one and pass another.
The decision tree
| Signal pattern | Action |
|---|---|
| Gets traffic, serves its intent well | Keep, monitor |
| Gets traffic, underperforms its intent | Improve |
| Two or more URLs serving the same intent | Consolidate |
| No traffic, has referring domains, a relevant destination exists | Redirect |
| No traffic, has referring domains, no relevant destination | Improve |
| No traffic, no links, a topic you still want to own | Improve or rewrite |
| No traffic, no links, nothing to save | Remove, or noindex if users still need the page |
Ask the questions in this order:
- Does it get traffic? If yes, keep or improve it.
- If not, do other sites link to it? If yes, redirect it to a relevant destination, or improve it if there is none. Either way, it isn’t a removal.
- If there are no links, would you still create this page today? If yes, improve or rewrite it.
- If not, can anything on it be salvaged into another page? Only when nothing can, remove it, or noindex it if users need it on the site.
The fourth question follows Google’s core updates guide, which calls deleting content a last resort, to be considered only when the content can’t be salvaged.
Consolidate by intent, not by keyword
Consolidate is an easy action to over-apply. Two pages that rank for the same query can still serve different readers, such as someone comparing options and someone following a how-to. Merging them removes the page that served one of those readers. The test is whether a reader of either page would be fully served by the other: if so, consolidate; if not, differentiate the two pages so each owns its job. Overlapping rankings call for a look, not a verdict.
When the answer is consolidate, merge before you redirect, so the surviving page carries what was useful on the others.
Redirect, remove and noindex are different tools
Redirect when the page holds links or history and a relevant destination exists. Relevance is part of the rule. Google’s site move guide warns against redirecting many old URLs to one irrelevant destination, such as the home page, and says this might be treated as a soft 404. A redirect to a page that doesn’t answer the same need isn’t a rescue.
Remove when there is nothing to keep and nowhere relevant to send the reader. For Google, 404 and 410 behave alike: its HTTP status codes documentation says all 4xx errors except 429 are treated the same, and indexed URLs that start returning a 4xx are removed from the index. Pick between them for accuracy, not for a ranking effect: a 410 says the page is gone on purpose, a 404 only that it wasn’t found.
Noindex when the page must stay live for visitors, such as a necessary utility page or a tag archive, but shouldn’t appear in search. The page keeps doing its job on the site and stays out of the results.
What removing pages can and can’t do
Google’s ranking systems guide says its systems are designed to work on the page level, with site-wide signals and classifiers also used, and that the helpful content system became part of the core ranking systems in March 2024. The same guide says that some poor site-wide signals don’t mean all the content from a site will rank poorly.
So an audit is not a quota of deletions. The core updates guide adds one case where deletion can help: if you are considering deleting whole sections, that is likely a sign they were made for search engines first, and deleting that unhelpful content can help the good content perform better. The audit’s job is to make that call page by page, with the joined data behind each verdict.
Roll out in batches and verify
Apply the verdicts in tracked batches and watch the affected URLs, so a wrong call means reversing one redirect instead of unwinding the whole audit. Record the reason for each verdict in the inventory, so the next audit can see why a page was kept or cut.
Two checks after each batch:
- Inspect the destination, not the old URL. Google’s help for the URL Inspection tool says the indexed result for a redirected URL reflects that URL, not the redirect target, and that the live test follows redirects without saying so. Inspect the keeper to see how Google handles the page that now carries the content.
- Expect removed URLs in the 404 list. The Page indexing report can list them as “Not found (404)”, and Google says 404 responses are not necessarily a problem if the page was removed without a replacement.
Frequently asked questions
Should I delete pages that get zero traffic?
Not on traffic alone. Check links first: a zero-traffic page with referring domains is a redirect candidate if a relevant destination exists, and an improve candidate if none does. Check intent next: if you would still create the page today, improve it. Remove only what has no traffic, no links and nothing salvageable.
Is a 410 better than a 404 for removed pages?
Not for Google. Its status codes documentation says all 4xx errors except 429 are treated the same. Choose the code that describes the removal accurately: 410 for a page removed on purpose, 404 for one that simply isn’t there.
How often should I run a content audit?
Tie it to change rather than a calendar. Re-run the inventory once you have published a meaningful volume of new pages, or after a core update changes how the site performs. Auditing too often re-decides pages that haven’t changed; auditing at long gaps lets pages that should have been improved or removed pile up unexamined.