Site Content Audit: What to Cut, Keep, and Fix
A site content audit is a systematic review of every page on your website, assessed against traffic, engagement, conversions, and strategic fit, to decide what to keep, improve, consolidate, or remove. Done properly, it is one of the highest-return activities in content strategy, because it forces you to treat your existing content as an asset base rather than an archive.
Most sites accumulate content the way offices accumulate paper. Nobody throws anything away, because discarding feels like waste. The result is a bloated, diluted site where good pages compete with weak ones for crawl budget, internal links, and reader attention. An audit breaks that cycle.
Key Takeaways
- A content audit is not a housekeeping exercise. It is a commercial decision about where your site’s authority and effort should concentrate.
- Most sites have more content than they need and less quality than they think. Cutting underperforming pages often lifts the performance of pages you keep.
- Traffic alone is a poor audit metric. A page that converts at 4% on low volume can outrank a high-traffic page that converts at 0.3%.
- Consolidation, not deletion, is usually the right move for thin content. Merging three weak pages into one strong one preserves equity and improves signal.
- An audit without a decision framework produces a spreadsheet, not a strategy. Build the criteria before you pull the data.
In This Article
- Why Most Sites Need an Audit More Than They Need New Content
- What Data Do You Actually Need Before You Start?
- How Do You Categorise Every Page?
- What Signals Actually Matter When Scoring Pages?
- How Do You Handle the Improve Category Without It Becoming a Backlog?
- What About Thin Content That Has Never Ranked?
- How Do You Turn Audit Findings Into a Repeatable System?
- What Are the Most Common Mistakes in a Content Audit?
Why Most Sites Need an Audit More Than They Need New Content
When I was running an agency and we took on a new SEO or content brief, the first question I asked was not “what should we create?” It was “what have you already got?” Nine times out of ten, the answer came back as a vague gesture toward a CMS with hundreds of entries and no clear picture of what was working. The instinct was always to produce more. The smarter move was almost always to start with what existed.
Content accumulates faster than it gets reviewed. A brand runs a campaign, publishes supporting blog posts, and moves on. A product gets discontinued but its page stays live. An SEO agency from three years ago produced 40 thin articles targeting long-tail terms that never ranked. None of it gets touched again, but all of it sits on the domain, diluting topical authority and consuming crawl budget that could be going to pages that actually matter.
Search engines are not impressed by volume. They are looking for consistent signals of quality, relevance, and authority. A site with 800 pages, 600 of which are thin, outdated, or redundant, sends weaker signals than a site with 200 pages that are all genuinely useful. That is a counterintuitive point for marketers who have been told that publishing more is always better. It is not.
If you want to think about content strategy at a structural level, the Content Strategy & Editorial hub covers the frameworks and decisions that sit behind this kind of work. An audit is one component of a broader editorial system, not a one-off tidy-up.
What Data Do You Actually Need Before You Start?
The temptation is to pull everything from Google Analytics and start sorting by sessions. That gives you a popularity contest, not an audit. Popularity is one signal. It is not the whole picture.
Before you open a single spreadsheet, you need to define what “good” looks like for your site. That means being clear on your content objectives: are you trying to rank for informational queries and build authority? Drive product page conversions? Support a sales team with mid-funnel material? The metrics that matter shift depending on the answer. A resource page for a B2B software company should be judged differently from a product category page for an e-commerce retailer.
With that framing in place, the data you need comes from four places. First, a crawl of your site using a tool like Screaming Frog or Sitebulb, which gives you a complete inventory of URLs, along with metadata, word count, indexation status, and technical signals. Second, Google Search Console, which tells you which pages are generating impressions and clicks from organic search, and which queries they are appearing for. Third, your analytics platform, whether that is GA4 or something else, for engagement metrics, conversion data, and assisted conversion paths. Fourth, your CMS, for publish dates and last-updated timestamps.
GA4 has changed how some of this data is structured, and Moz has written clearly about using GA4 data to inform content strategy decisions, which is worth reading if you are still finding your feet with the new reporting model. The core principle has not changed: you are trying to understand which pages are earning their place and which are not.
One thing I would add from experience: do not skip the qualitative pass. Metrics tell you what is happening. They do not always tell you why. A page with high traffic and low time-on-page might be failing because the content is thin, or because it is answering the question so quickly that users get what they need and leave. Those are very different problems requiring very different responses.
How Do You Categorise Every Page?
The standard audit framework puts every page into one of four buckets: keep, improve, consolidate, or remove. That is the right structure. The difficult part is applying it consistently, especially when you have hundreds or thousands of URLs to work through.
Keep means the page is performing well against your defined criteria, is technically sound, and does not need significant intervention. These pages still need monitoring, but they are not your priority.
Improve means the page has clear potential but is underperforming. This might be a page that ranks on page two for a valuable query and needs stronger content, better internal linking, or a more compelling title tag. It might be a page with good traffic but weak conversion, where the call to action or the content structure needs rethinking. Improvement work is where most of your post-audit effort should go, because you are working with existing equity rather than starting from zero.
Consolidate means you have multiple pages covering the same topic or query at a level of depth that none of them can rank effectively individually. The answer is to merge them into one authoritative piece, redirect the others, and concentrate the signals. I have seen this move lift organic traffic on merged pages by a meaningful margin, not because of any trick, but because the combined page is simply better and more complete than any of the originals were separately.
Remove means the page has no traffic, no backlinks, no strategic value, and no realistic path to any of those things. Remove it, redirect it if there is any residual equity to preserve, and stop worrying about it. The reluctance to delete is understandable but usually misplaced. A page that has generated zero organic sessions in 18 months is not an asset. It is noise.
The Content Marketing Institute has a useful framing for how content decisions connect back to audience needs, and their work on understanding your target audience within a content framework is a sensible anchor when you are trying to decide whether a piece of content has a future or not.
What Signals Actually Matter When Scoring Pages?
I want to be specific here, because vague scoring criteria produce vague decisions. When I have run audits, either for agencies I led or for clients, the signals I weight most heavily are these.
Organic traffic and trend direction. Not just current volume, but whether it is growing, flat, or declining. A page with modest traffic that has doubled in six months deserves a different decision than a page with higher traffic that has halved.
Conversion contribution. This includes direct conversions and assisted conversions. A page that rarely converts directly but frequently appears in the path to conversion is doing useful work. Strip it out and you may not see the impact immediately, but you will feel it.
Keyword ranking position and query intent alignment. A page ranking 15th for a query it was designed to rank for is a candidate for improvement. A page ranking 15th for a query it was never designed to target is a candidate for a different kind of attention, because it might be telling you something about what your audience actually wants.
Backlink profile. Pages with external links pointing to them carry equity. You do not delete those pages without a redirect strategy. Even if the content is outdated, the links have value that needs to be preserved or transferred.
Content quality on a read-through. This is subjective, but it matters. Is the page genuinely useful? Does it answer the question it is supposed to answer? Is it better than what ranks above it, or is it thinner? Metrics can obscure quality problems. A human read is still part of the process.
Copyblogger has written about the relationship between SEO and content quality in a way that holds up, and the core argument, that content which genuinely serves readers tends to perform better over time, is as true now as it was when they first made it.
How Do You Handle the Improve Category Without It Becoming a Backlog?
This is where most audits fall apart. The keep and remove decisions are relatively clean. The improve bucket is where good intentions go to die, because it requires ongoing editorial work rather than a one-time decision.
The discipline I have found most useful is triage by potential return. Not every page in the improve bucket deserves equal attention. Prioritise the pages where the gap between current performance and realistic potential is largest, and where the effort required to close that gap is manageable. A page ranking 11th for a term with strong commercial intent is a higher priority than a page ranking 40th for a term with no conversion history.
Set a realistic improvement schedule. If you have 60 pages in the improve bucket and a content team of two people, you are not going to fix all of them in a quarter. Pick the top 15, assign owners, set deadlines, and treat it like a project rather than a wish list.
Also think about what “improve” actually means for each page. Sometimes it is a structural rewrite. Sometimes it is adding a section that addresses a related query the page is already getting impressions for. Sometimes it is updating statistics and examples that have aged. Sometimes it is fixing the internal linking so that the page receives more authority from the rest of the site. These are different tasks with different time requirements, and conflating them leads to underestimating the work.
HubSpot has a useful overview of content distribution strategy that is worth reading alongside improvement planning, because improved content that does not get redistributed often fails to recover its rankings as quickly as it should. Updating the page is only half the work.
What About Thin Content That Has Never Ranked?
Thin content is the category that generates the most debate in audits. Marketers get attached to content they produced, even when the data is clear that it has never done anything useful. I have been in rooms where someone has argued for keeping a 300-word blog post from 2019 because “it might still rank eventually.” It will not. And it is actively dragging down the pages around it.
The honest test for thin content is this: if a prospective customer landed on this page with no other context about your brand, would it make them more or less likely to trust you? If the answer is less, or even neutral, the page is not earning its place.
The options for thin content are consolidation or removal. If the topic is genuinely relevant to your audience and your strategy, consolidate the thin page into a stronger piece that covers the subject properly. If the topic is not relevant, or the search volume is negligible, remove it and redirect to the closest relevant page.
One thing I would caution against is using AI to bulk-expand thin content as a shortcut. I have seen this done, and the result is usually pages that are longer but not better. Moz has written about scaling content with AI in a way that is sensible about this distinction: volume is not the same as value, and AI-assisted content still requires editorial judgment to be genuinely useful.
How Do You Turn Audit Findings Into a Repeatable System?
A one-time audit is better than nothing. A repeatable audit process is significantly better than a one-time audit. The goal is to build content review into your editorial calendar rather than treating it as a crisis response every two years when you notice organic traffic has dropped.
The practical structure I have seen work is a rolling quarterly review of your top 20% of pages by traffic and conversion, combined with an annual full-site audit. The quarterly review catches decay early, when a page drops from position 3 to position 9, before it falls off page one entirely. The annual audit handles the structural decisions: consolidations, large-scale removals, and strategic repositioning.
Build a living audit document rather than a static spreadsheet. Every page should have a status, a decision, an owner, and a next-review date. When new content is published, it goes into the document from day one. When content is updated, the date and the nature of the update are logged. Over time, this becomes an editorial record that tells you not just where your content is, but how it has evolved and why decisions were made.
The Content Marketing Institute’s work on content marketing frameworks and process is a useful reference for how auditing connects to broader editorial governance. An audit in isolation is a project. An audit embedded in a process is a system.
If you want to think about how a content audit fits into a wider editorial strategy, including how to build the governance structures that make this kind of work sustainable, the Content Strategy & Editorial hub covers the full picture across planning, production, and measurement.
What Are the Most Common Mistakes in a Content Audit?
The first mistake is auditing without a strategy. If you do not know what your content is supposed to do commercially, you cannot make good decisions about which pages are working and which are not. An audit is a strategic exercise that requires strategic context. Pulling data without that context produces a spreadsheet full of numbers and no clear direction.
The second mistake is using traffic as the only metric. I have seen teams preserve high-traffic pages that were generating zero conversions and zero brand value, simply because the session numbers looked good. Traffic is a means to an end. The end is business outcome. Keep your eye on what actually matters.
The third mistake is deleting pages without redirects. Every time I have audited a site that has done previous content clean-ups, I find a graveyard of 404s where pages were removed without redirect strategies. Those pages often had backlinks, internal links, or direct traffic. The equity evaporates when the page disappears without a redirect. Always map your removals to redirect destinations before you touch anything.
The fourth mistake is treating the audit as a one-person job. The content team can pull the data and make initial recommendations, but the decisions about what to keep, cut, or invest in often require input from sales, product, and commercial leadership. Content that looks weak from an SEO perspective might be critical for the sales team. Content that looks strong in traffic terms might be attracting the wrong audience entirely. Those conversations need to happen before decisions are finalised.
The fifth mistake is auditing and then doing nothing. I have seen this more times than I care to count. A thorough audit produces a clear action list, and then it sits in a shared drive for six months while everyone gets back to producing new content. The audit was useful work. The failure to act on it made it pointless. Build the implementation plan before you finish the audit, not after.
About the Author
Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.
