Quick answer
A content audit lists every indexable page on your site, pulls the same metrics for each and gives every URL one decision: keep, improve, merge or remove. Build the inventory from a crawl, Search Console and analytics, score each page, then handle redirects safely when merging or removing, and schedule refreshes so the audit leads to action.
Ahrefs (2023) found that 96.55 per cent of the roughly 14 billion pages in its index get zero organic traffic from Google, which is why auditing weak pages is worthwhile.
A content audit means listing every indexable page on your site, pulling the same set of metrics for each, and giving every URL one decision: keep, improve, merge or remove. In this guide you will build the inventory from a crawl, Search Console and analytics, learn which metrics to pull and how to score them, handle redirects safely when you merge or remove pages, and set a refresh schedule so the audit does not have to start from zero next year.
When a content audit is worth doing
Sites accumulate pages. Old news posts, thin tag archives, product pages for items you no longer stock, three blog posts on the same subject written by three different people. None of these look harmful on their own, but together they dilute your site’s focus and waste the time search engines spend crawling it.
A content audit is worth the effort when:
- Your site has more pages than you can name from memory, and nobody owns the older ones.
- Organic traffic has plateaued or slipped despite new content going live.
- You are planning a redesign or migration and want to avoid moving dead weight.
- A broader audit has flagged duplicate, thin or competing pages.
If your traffic fell sharply rather than slowly, run through a sudden traffic drop checklist first. A content audit is a deliberate clean-up, not an emergency response.
Step 1: Build the page inventory from three sources
No single tool sees every page. Combine three sources into one spreadsheet, with one row per URL.
- A site crawl. Use any desktop or cloud crawler to collect every URL reachable through internal links, with its status code, indexability, canonical, title, word count and number of internal links pointing to it.
- Search Console. Export pages with impressions and clicks over a long period, ideally twelve months to smooth out seasonality. This surfaces URLs that Google knows about but your crawler did not reach, such as orphan pages.
- Analytics. Export landing pages with organic sessions and conversions (or key events) over the same period.
Optionally, add your XML sitemap and a backlink export so you can see which pages have external links. Merge everything on the URL column. Watch for small mismatches such as trailing slashes, uppercase letters and tracking parameters; normalise these before merging or you will end up with duplicate rows.
Filter out non-indexable URLs that are meant to be that way, like cart pages and internal search results. Keep a note of them, but they do not need a content decision.
Step 2: The metrics to pull for every page
Keep the column set focused. More columns do not make better decisions. These are the ones I use:
| Metric | Source | What it tells you |
|---|---|---|
| Clicks and impressions (12 months) | Search Console | Whether the page has search demand and whether it captures it |
| Top query and average position | Search Console | What Google thinks the page is about |
| Organic sessions and conversions | Analytics | Business value |
| Referring domains | Backlink tool | Whether removing the page would waste earned links |
| Internal links in | Crawl | How much the site itself supports the page |
| Word count and last updated date | Crawl or CMS | Rough signals of depth and freshness |
| Topic or cluster | Manual | Which pages overlap and belong together |
| Content quality notes | Manual review | Accuracy, intent match, usefulness |
Word count is a signal, not a target. A short page that answers a quick question perfectly is fine. Use it to spot pages that may be too thin for the topic they cover, then judge them by reading.
Step 3: Score each page
A simple scoring system speeds up decisions on large sites. Give each page a score of 0 to 2 on four questions:
- Demand: Does the page get meaningful impressions for relevant queries? (0 none, 1 some, 2 strong)
- Value: Does it bring sessions that convert or support conversions? (0 none, 1 indirect, 2 direct)
- Authority: Does it have referring domains or strong internal links? (0 none, 1 some, 2 strong)
- Quality: Is it accurate, current and a good match for search intent? (0 poor, 1 acceptable, 2 strong)
The total is out of 8. The score does not make the decision for you. It sorts the spreadsheet so you review the obvious keeps and obvious removals quickly, and spend your reading time on the middle.
Step 4: Decide keep, improve, merge or remove
Every row gets exactly one action. Here is how I decide.
Keep
The page scores well on quality and either demand or value. It needs no work beyond routine checks. Most of your best pages land here.
Improve
The page has demand or authority but weak quality or a poor intent match. Typical work: rewrite the introduction to answer the query sooner, update outdated facts, add missing sections that competing pages cover, improve the title, and add internal links from related pages.
Merge
Two or more pages cover the same topic and split the signals between them. Choose the strongest URL as the destination (usually the one with the most links, clicks or the cleanest URL), fold the useful parts of the others into it, then redirect the others. If you are unsure whether pages genuinely overlap, the method in my guide to keyword cannibalisation audits helps you confirm it from query data.
Remove
The page has no demand, no value, no links and no realistic route to improvement. Expired event pages, empty tag archives and thin duplicates often fall here. Removal does not always mean deletion: some pages are useful to users but not to search, and can simply be set to noindex.
Here is an illustrative example. Imagine a travel agency in Dubai with forty blog posts about visa requirements, many written years apart. A content audit might keep the five posts with steady traffic, improve eight that still earn impressions but contain outdated rules, merge eighteen overlapping posts into four country guides, and remove nine announcements about expired promotions.
Step 5: Handle redirects when merging and removing
Redirect mistakes undo the benefit of the audit. Follow these rules:
- Use a single permanent (301) redirect from each merged URL directly to its final destination. Avoid chains.
- Redirect to the most relevant equivalent page, not the homepage. Redirecting unrelated pages to the homepage is often treated by Google as a soft 404 anyway.
- For removed pages with no relevant equivalent and no links, returning a 404 or 410 is acceptable. For removed pages with valuable links, find the closest relevant page and redirect there.
- Update internal links to point at the destination URLs, so you are not relying on redirects inside your own site.
- Remove redirected and deleted URLs from your XML sitemap.
- Keep a redirect map in the spreadsheet: old URL, new URL, date, reason. Your future self will thank you during the next migration.
After the changes go live, crawl the old URLs to confirm each one lands where intended, and watch Search Console’s indexing reports over the following weeks.
Step 6: Turn decisions into work and schedule refreshes
An audit of a large site can produce hundreds of actions. Rank them before starting; the method in my post on ranking SEO audit findings by impact and effort applies just as well to content work. As a rule, merges involving pages with traffic go first, then improvements to pages close to ranking well, then removals.
To stop the site drifting back into disorder, add two columns to your spreadsheet: owner and next review date. Then set review cycles by page type:
- Revenue and service pages: review regularly, and whenever your offer or pricing changes.
- Guides on topics where facts change, such as regulations, prices or software: review on a fixed schedule.
- Evergreen explainers: review less often, or when their traffic declines.
- News and announcements: decide at publication whether they will be updated, archived or removed later.
For a wider view of site health beyond content, my SEO checklist covers the other areas to keep an eye on between audits.
Questions people ask about content audits
Will deleting pages hurt my rankings?
Removing pages with no traffic, links or value rarely causes harm and can help search engines focus on your better pages. The risk comes from removing pages that do have links or traffic without redirecting them. That is why the inventory includes those metrics.
Should I noindex or delete thin pages?
Noindex pages that users still need, such as a policy page or a login page. Delete or redirect pages that serve no one. Noindexed pages still get crawled, so a site full of them still carries some crawling cost.
How many pages can I review by hand?
Fewer than you think. Use the scoring to batch the clear decisions, then read the pages in the uncertain middle. On very large sites, review templates and samples rather than every URL.
Is a content audit part of a full SEO audit?
It should be. Content is one of the core areas, alongside technical health, links, local presence and tracking. My overview of what a professional SEO audit includes explains how they fit together.
If your inventory is already larger than you can face, I run content audits as part of my professional SEO audit services, with a decision for every URL and a redirect map ready for your developer. Request a free audit and I will tell you whether a content audit is the right next step for your site.