How to Do a Content Audit (Step by Step)

To do a content audit, list every published page in a spreadsheet, record how each one performs (position, clicks, impressions, and whether traffic is rising or falling), check each page against the live search results for its target keyword, and assign one decision per page: keep, refresh, consolidate, or prune. The output is a prioritized list of fixes, ordered by how much traffic each page can realistically recover. That is the whole method, and the rest of this page is how to run each step without it turning into a month-long project.

Why bother with a content audit at all

Most sites publish for years and audit never. The result is a library where a quarter of the pages do real work, a quarter used to and quietly stopped, and half never ranked for anything. You cannot see which is which from the CMS, because the dashboard shows what you published, not what is performing. A content audit is the pass that separates the three groups so your next month of effort goes to the pages that can pay it back.

The reason this is the cheapest growth available is that the content cost is already sunk. A page that ranked in 2022 and slipped in 2025 still has its URL age, its backlinks, and its ranking history. Refreshing it competes from a running start, which is why refreshed pages tend to move in weeks while brand-new pages take months. Publishing more is the instinct; auditing what you have is usually the better trade.

Step 1: Inventory every URL

Start with a complete list of your indexable pages, one per row. Pull it from your XML sitemap, a crawl with a tool like Screaming Frog, or a CMS export. For each URL, capture the basics you will not want to look up twice: the page title, the publish date, the last-modified date, and the word count. This is bookkeeping, not judgment, so move fast. The goal of step one is simply to know how many pages you actually have, which is almost always more than the team guesses.

If the site is large, scope the audit to one section at a time (the blog, then the resource center, then the product pages) rather than trying to boil the whole domain in one sitting. A finished audit of 80 blog posts beats an abandoned audit of 800 mixed URLs.

Step 2: Pull the performance data

Now add the numbers that turn an inventory into a decision tool. For each URL, record from Google Search Console its top query, average position, clicks, and impressions over the last three months. Then do the one comparison most audits skip: pull the same metrics for the same period a year earlier and note the direction of travel. A page holding 400 clicks a month is a different situation from a page that did 1,200 last year and does 400 now, even though today's number is identical.

Add a conversion or goal column if you can, even a rough one. A page can rank and get traffic and still be worthless if the visitors it brings never do anything. Judging a page on rankings alone is how teams end up lovingly refreshing content that has never once produced a lead. Part of that judgment is external too: watching what competitors publish on the same topics and how audiences respond tells you whether the whole subject is heating up or cooling off, and a tool that tracks competitor content and audience reaction across the web makes that a five-minute check instead of a manual trawl.

Step 3: Judge intent against the live SERP

This is the step that separates a real content audit from a spreadsheet exercise. For each page that matters, open the current top five results for its target keyword and read them like an editor. What question are they answering? In what format, a step-by-step guide, a comparison table, a definition, a tool? In what order? That live result set is the bar your page is actually being measured against, and it moves over time. A page written in 2021 to answer "what is X" can lose simply because the query now wants "how to do X," and no amount of on-page polish fixes a page that answers the wrong question.

Mark each page yes, partly, or no on intent match. The "no" rows are not refresh candidates; they are rewrite-the-angle candidates or prune candidates, and treating them as quick refreshes is how audits waste time. The "partly" rows, pages that answer the right question but miss subtopics the ranking pages all cover, are usually your best recoverable traffic.

Step 4: Decide keep, refresh, consolidate, or prune

Every row lands in exactly one bucket. Here is the rule set.

• Keep when a page ranks well, matches intent, and holds or grows its traffic. Leave it alone and re-check next quarter. Do not fix what is working.
• Refresh when a page has history and links but its traffic is sliding, or it sits on page two with a gap you can name. This is where the money is. Rewrite the title, close the subtopic gaps, update stale facts, add internal links, and ship the schema it qualifies for.
• Consolidate when two or more thin pages compete for the same intent. Merge them into the single strongest URL and 301-redirect the others to it. One authoritative page beats three overlapping weak ones, and the redirects pass the equity forward.
• Prune only when a page has no traffic, no links, no conversions, and no strategic reason to exist. Redirect it to a relevant page if it retains any equity; otherwise remove it. Prune sparingly, because the instinct to delete after an audit is almost always too aggressive.

That last warning is worth repeating. Pruning feels like progress because deleting is visible and satisfying, but a page with even a handful of backlinks and some ranking history is usually worth a refresh or a merge before it is worth a delete. Rebuilding that history from zero is far more expensive than editing what you already have.

Step 5: Prioritize and execute

You now have a list of refresh, consolidate, and prune actions. Do not work it top to bottom by URL. Order it by traffic at stake: a page ranking eight for a keyword with real search volume is a bigger prize than a page ranking forty for a term nobody searches, even though the second looks worse. A quick proxy is impressions multiplied by the size of the position gap. Fix the pages where a small ranking move captures the most traffic first.

Then execute in batches. Refreshes are editorial work: run each page through the same on-page checklist so the work is consistent and nothing gets missed. If you want the exact spreadsheet columns and the per-page checklist laid out to copy, the content audit template has both, and the content audit tool automates the diff against the live SERP so you spend your time editing instead of comparing browser tabs.

How long should a content audit take?

For a library of 50 to 100 pages, a thorough manual audit is roughly a focused week: a day to inventory and pull data, two to three days to judge intent and assign decisions, and the rest to start executing. The judgment steps are the slow part, not the data collection. This is exactly why teams put off auditing until traffic has already fallen, and why automating the repetitive diff work matters on larger sites, where a page-by-page manual pass simply does not scale and decay never stops accumulating in the meantime.

Frequently asked questions

How do you perform a content audit?

Perform a content audit in five steps: inventory every URL in a spreadsheet, pull each page's performance data from Search Console including the year-over-year trend, judge whether each page still matches the intent of the live search results for its keyword, assign one decision per page (keep, refresh, consolidate, or prune), then prioritize the refresh and consolidate work by how much traffic each page can recover.

What tools do I need for a content audit?

At minimum you need Google Search Console for performance data, a crawler or CMS export to inventory URLs, and a spreadsheet to record it all. Analytics adds conversion data, which is worth including so you do not refresh pages that get traffic but never convert. Dedicated audit tools automate the diff between your page and the pages ranking above it, which is the most time-consuming manual step.

How is a content audit different from an SEO audit?

A content audit evaluates the quality, relevance, and performance of individual pages and decides what to do with each one. A technical SEO audit evaluates the whole site's crawlability, speed, indexation, and structure. They overlap but answer different questions: the content audit asks which pages to keep and improve, while the technical audit asks whether search engines can access and understand the site at all.

How often should you audit your content?

Run a full content audit once a year, with a lighter decay check each quarter that only reviews the pages losing the most traffic. Large or fast-publishing sites benefit from continuous auditing rather than an annual event, because content decay is gradual and constant, so a scheduled re-scan keeps the refresh list current instead of letting a year of quiet decline pile up before anyone looks.