A content audit is the least exciting SEO task there is, and one of the most useful. You go through every page on your site, check how each one is doing, and decide what to keep, what to improve, what to merge and what to remove.
I recently did a full audit of this site, nearly 300 posts and pages, most of them written between 2021 and 2023. I expected to find a few outdated articles. I found broken images in dozens of posts, two articles competing for the same keyword, links pointing to websites that had nothing to do with the text, and the same promotional paragraphs pasted into a dozen different reports. None of it was visible from the dashboard.
This guide is the process I used, with the real problems it uncovered, so you can run the same audit on your own blog.
What is a content audit?
A content audit is a page-by-page review of the content on a website, using performance data and a manual check of quality, to decide what should happen to each page. The output is not a report: it is a list of actions, one per URL.
It is different from a technical SEO audit, although they overlap. A technical audit asks “can Google crawl and index this site?”. A content audit asks “is each page on this site still worth reading, and is it the best page we have on its topic?”
When you should do one
- Your traffic has dropped and you cannot point to a single page that explains it.
- The site is more than two or three years old and nobody has reread the early posts.
- Several people have written for the site, or you bought it from someone else.
- You migrated hosts, themes or domains, or ran bulk find-and-replace operations in the database.
- You plan to publish a lot of new content and want to know what already exists first.
For a small blog, once a year is enough. For a site that publishes weekly, a light quarterly pass on the pages that lost the most clicks works well.
What Google says about removing and updating content
Before touching anything, it helps to know what does not work. Google’s own guide to creating helpful, people-first content lists two practices as warning signs: changing the date on pages that have not substantially changed, and adding or removing lots of content mainly to make the site seem “fresh”. In its words, that will not help.
When CNET deleted thousands of old articles in 2023 to improve its rankings, Google’s Search Liaison, Danny Sullivan, responded publicly that the idea of Google disliking old content is “not a thing”, and that older content can still be helpful (reported by Gizmodo).
So the goal of an audit is not to delete, and not to change dates. It is to make every page that stays genuinely useful, and to deal honestly with the pages that are not.
How to do a content audit, step by step
Step 1: Build a list of every URL
Start with a complete inventory. On WordPress you can export all posts and pages, but I prefer a crawl because it also catches what the CMS hides: redirects, old tag pages, attachment pages. Screaming Frog SEO Spider is free for up to 500 URLs, which covers most small blogs.
Add the URLs from your XML sitemap too. Anything in the sitemap that the crawler did not find is a page with no internal links pointing to it (an orphan page).
Step 2: Add performance data
Export the Pages report from Google Search Console for the last 12 months (Search Console keeps 16 months, so you can also compare with the year before). For each URL you want at least:
| Column | Where it comes from | Why it matters |
|---|---|---|
| Clicks and impressions | Search Console | Is anyone finding this page? |
| Top query and average position | Search Console | What Google thinks the page is about |
| Change vs previous year | Search Console | Spots decaying pages early |
| Referring domains | Any backlink tool | Pages with links need care before removing |
| Word count and last real update | Crawl / CMS | Flags thin and stale pages |
| Conversions or revenue | Analytics, affiliate reports | Some low-traffic pages still earn |
If you want to go further, my guide to measuring niche blog performance covers which metrics are worth tracking over time.
Step 3: Read the pages (yes, all of them)
This is the step everyone wants to skip, and it is where I found almost every serious problem. Data tells you a page gets no clicks; only reading it tells you why. For each page, I check:
- Is the information still correct? Prices, tools, statistics and screenshots age fast.
- Does every link go where the text says it goes?
- Do all images load?
- Is there text that does not belong to this topic (boilerplate, off-topic promotions)?
- Is there another page on the site that answers the same question?
- Would I be happy to send this page to a reader today?
Step 4: Give every page one action
Each URL gets exactly one decision. These are the six I use:
| Action | When | What to do |
|---|---|---|
| Keep | Accurate, useful, performing | Nothing, or small fixes |
| Update | Good topic, outdated or incomplete content | Rewrite what aged, fix links and images, date it honestly |
| Merge | Two or more pages answer the same question | Combine the best of both into one, 301 the others |
| Redirect | Page has no value but has links or traffic | 301 to the closest equivalent page |
| Noindex | Useful to visitors but not for search (thank-you pages, filters) | Keep it, remove it from search and sitemap |
| Remove | No value, no traffic, no links, no equivalent | Delete and let it return 404/410 |
Two rules keep this honest. First, only redirect to a page that really covers the same thing; Google’s documentation on redirects treats a 301 as “this content has moved”, not as a way to pass authority to something unrelated. Second, “remove” is the last option, not the first. Most pages deserve an update before they deserve a delete.
Step 5: Prioritize and fix
You will end up with more actions than time. I sort them in this order:
- Broken things (images, links, redirects): fast to fix and they hurt every visitor.
- Merges: two weak pages competing for one keyword often become one page that ranks.
- Updates to pages close to page one (positions 8-30 with real search volume).
- Everything else, a few pages per week.
Step 6: Measure again in two to three months
Write down the date of every change and the position before it. Rankings move slowly after an update, so check the same pages again after eight to twelve weeks. If a page you rewrote has not moved at all, the problem is probably not on the page: it may be the search intent, the competition, or the site as a whole.
What my own content audit found
Here is what came up when I ran this process on this site. I am sharing it because these problems are common on any blog that is a few years old, and almost none of them show up in a dashboard.
1. Two articles targeting the same keyword
I had one post on “niche marketplace” and another on “niche marketplaces”, written at different times and covering the same ground. Neither ranked well. I merged the best parts into a single, rewritten guide on niche marketplaces, redirected the other URL to it and updated the five internal links that pointed to the old one.
2. Hundreds of broken image paths from one bulk edit
At some point a bulk find-and-replace across the database changed “2024” to “2025” to refresh the year in the articles. It also changed the year inside image URLs, which point to upload folders named by date. The result: 490 image paths in 75 posts pointed to folders that did not exist. The posts looked fine in the editor list; you only saw the broken images by opening them. Restoring the original paths fixed them.
3. Images that never existed
Almost 200 more image files referenced in old posts were not on the server at all, and not in the backups either: they had been lost in an earlier migration. I found 12 of them elsewhere on the server under a different folder and fixed those. The remaining broken images were removed from the posts (with their empty captions) so readers no longer see broken image icons.
4. Links that went to the wrong place
This one surprised me the most. In one industry report, a link labeled as a pet retailer went to a fitness coaching page, and another labeled as a pet-sitting company went to a boxing gym. A market statistic was credited to one source but linked to a press release about the sports industry. In another report I found a hidden tracking pixel and a link to an unrelated local business. Nobody clicks these links to check them, which is exactly why they survive for years.
5. The same promotional paragraphs everywhere
About a dozen of my niche reports (for example the gaming and pet reports) contained the same paragraphs recommending the same three tools, whatever the niche. It is the kind of copy-and-paste text that makes a page feel written for search engines rather than for the reader. I am rewriting those reports one by one, replacing the boilerplate with information specific to each niche.
The common thread: every problem came from something done in bulk, or from content written once and never reread. If you are building a site now, the cheapest audit is the one you avoid by publishing fewer, better pages. I wrote more about that in my guide to topical authority.
Content audit checklist
A short version you can copy into your spreadsheet:
- Every URL from the crawl and the sitemap is listed.
- Clicks, impressions, top query and position added for each URL.
- Referring domains added (never delete a page with links without redirecting it).
- Each page read in full, with notes on outdated facts, broken links and images.
- Duplicate topics identified and grouped.
- One action per URL: keep, update, merge, redirect, noindex or remove.
- Internal links updated after every merge or redirect.
- Date of each change recorded, and a reminder set to check again in 8-12 weeks.
Content audit FAQ
How long does a content audit take?
Collecting the data takes an afternoon. Reading the pages is what takes time: for a blog with 100-300 posts, plan for a few days of reading spread over a couple of weeks. It is worth splitting by category so you can act on one section before moving to the next.
Should I delete old blog posts?
Not because they are old. Delete a page only when it has no useful content, no traffic, no links and no better page it could be redirected to. Old posts that are still accurate can keep ranking for years.
Does updating the date help rankings?
Changing the date without changing the content does not help, and Google lists it as a warning sign. Update the date when you have made a real update, and it is good practice to say briefly what changed.
What is the difference between a content audit and a technical SEO audit?
A technical audit checks whether search engines can crawl, render and index your pages. A content audit checks whether the pages themselves are accurate, useful and not competing with each other. You need both, and a content audit often uncovers technical issues along the way, like the broken images I found.
Can I do a content audit without paid tools?
Yes. Google Search Console, a free crawler limit of 500 URLs, a spreadsheet and your own time are enough for most niche sites. Paid tools help with backlink data and larger sites.
For more on growing a blog after the cleanup, see my guides on how to grow a niche blog, the most common blogging mistakes and the niche blogging hub.