Web Content Audit: What to Keep, Cut, and Fix

A web content audit is a systematic review of every page on your site, assessed against performance data, business goals, and audience relevance, to decide what to keep, improve, consolidate, or remove. Done properly, it is one of the highest-return activities in content strategy, because it works on what you already have rather than asking you to produce more.

Most sites accumulate content the way organisations accumulate processes: gradually, without a clear owner, and with no one ever stopping to ask whether it is still earning its place. A content audit forces that question across every URL you own.

Why Most Sites Are Carrying Dead Weight

When I ran a performance marketing agency, we inherited client sites that had been publishing content for five or six years without any systematic review. The pattern was almost always the same: a core of genuinely useful, well-trafficked pages surrounded by a much larger volume of pages that were pulling negligible organic traffic, had no backlinks, and were converting nobody. Some of those pages were actively competing with stronger pages for the same search terms.

The instinct, when you see that, is to want to fix everything. But that is the wrong frame. The right question is: what is this content costing you? Crawl budget, internal link equity, topical dilution, and the time your team spends maintaining content that does nothing for the business. These are real costs, even if they do not show up as a line item on a P&L.

Content strategy is broader than any single audit, and if you want context for how audits fit into a wider editorial programme, the Content Strategy & Editorial hub covers the full picture. But the audit is where most programmes should start, because you cannot build a coherent strategy on top of a site you do not fully understand.

What Does a Content Audit Actually Cover?

A thorough audit covers four dimensions: performance data, content quality, technical health, and strategic fit. Most audits that go wrong focus almost entirely on the first and ignore the other three.

Performance data includes organic traffic, rankings, click-through rate, time on page, bounce rate, and conversion events. You pull this from Google Search Console, Google Analytics, and your rank tracker. The Semrush content audit guide covers the tooling side of this well if you want a practical walkthrough of how to pull the data together.

Content quality is a manual assessment. Does the page answer the question it is targeting? Is the information accurate and current? Is the depth appropriate for the intent? Does it reflect the expertise and perspective your brand actually has, or is it generic filler that could have been published by anyone?

Technical health covers page speed, mobile rendering, indexation status, canonical tags, and internal linking. A page can be well-written and still underperform because it has a canonical pointing somewhere else or because it sits three clicks from the homepage with no internal links pointing to it.

Strategic fit is the question most audits skip entirely. Does this page support a product or service you are actually selling? Does it attract the audience you want? Is it consistent with where the business is going, or does it reflect a positioning you moved away from two years ago? I have seen sites carrying entire content categories built around services the company no longer offered. Nobody had ever stopped to remove them because nobody owned the decision.

How to Build Your Content Inventory

Before you can audit anything, you need a complete list of what you have. Use a crawler, Screaming Frog is the standard choice, to pull every indexable URL on the site. Export that to a spreadsheet and add columns for your key performance metrics.

For each URL, you want at minimum: organic sessions over the last twelve months, organic impressions, average position for the primary keyword, number of referring domains, and the page’s primary conversion event if one exists. If you have a site with fewer than 500 pages, you can also add a manual quality rating column. For larger sites, you will need to triage by traffic threshold first and then manually review the pages that fall into grey zones.

One thing I have learned from doing this across a wide range of industries: the inventory itself is often a revelation for clients. Many marketing teams genuinely do not know how many pages their site has, and when they see it listed out, the volume of low-quality content is usually a shock. I have seen B2B sites with under 200 products carrying 1,400 blog posts, most of which had never received a single organic visit.

The Four Decisions: Keep, Improve, Consolidate, Remove

Every page in your inventory needs to end up in one of four buckets. The discipline is in applying consistent criteria rather than making case-by-case judgements based on gut feel.

Keep means the page is performing well, is strategically relevant, and needs no significant work. These pages still need monitoring, but they are not your priority. Be honest about how few pages actually qualify here. Most sites have fewer high-performing pages than their owners think.

Improve means the page has strategic value or ranking potential but is underperforming relative to what it could do. This might be a page targeting a high-value keyword that ranks on page two, or a page with good traffic but a weak conversion rate. Improvement work is prioritised by the gap between current and potential performance, not by how easy the fix is.

Consolidate means you have multiple pages covering the same or overlapping topics, and combining them into a single, stronger page would serve both users and search engines better. Keyword cannibalisation is the usual trigger here. When two of your own pages are competing for the same search term, neither ranks as well as one authoritative page would. The Moz piece on adjusting content strategy is worth reading for context on how search behaviour is shifting and why consolidation is becoming more important, not less.

Remove is the decision most teams avoid, and it is usually the one with the highest return. Pages that have received no organic traffic in twelve months, have no backlinks, target no keyword of commercial relevance, and add nothing to the user experience are not neutral. They dilute your site’s topical authority and waste crawl budget. Remove them, redirect if there is any residual link equity, and move on.

The hardest calls are the ones where someone in the business is attached to the content. I have sat in rooms where a piece of content was effectively being defended because a senior person wrote it three years ago. That is not a content strategy conversation, it is a political one. The answer is to make the decision criteria explicit before you start the audit, so the framework does the work rather than the conversation becoming personal.

Where AI Changes the Audit Calculation

The rise of AI-generated content has changed the context in which audits happen. Sites that published large volumes of AI content without sufficient editorial oversight are now carrying pages that are technically indexed but functionally invisible. Search engines have become better at identifying content that lacks genuine expertise, and the bar for what constitutes a useful page has risen as a result.

If your site published AI content at scale over the last two years, your audit needs to include a quality assessment that goes beyond performance metrics. A page can have reasonable traffic from long-tail queries while still being the kind of thin, generic content that erodes trust with the readers who do land on it. The Moz overview of AI content creation is a useful reference point for understanding where the quality floor currently sits.

The audit question for AI content is not whether it was written by a machine. It is whether it demonstrates the kind of specific, grounded expertise that a reader could not get from a generic source. Content that could have been written by anyone, about anything, for anyone, is not earning its place on your site regardless of how it was produced.

Understanding your audience deeply enough to make those quality calls is a prerequisite. The Content Marketing Institute’s framework for defining your target audience is worth revisiting if you are not confident that your current content reflects a clear picture of who you are actually writing for.

How to Prioritise the Action Plan

An audit that produces a spreadsheet with 800 rows and no clear sequence is not useful. The output needs to be a prioritised action plan that your team can actually execute against.

I prioritise audit actions using three criteria: commercial impact, effort required, and risk of inaction. A page ranking on position eleven for a term that drives qualified leads should be at the top of the improve list regardless of how much work it needs. A page with no traffic, no links, and no strategic relevance should be removed this week, not in Q3.

For consolidation work, sequence by the pages that are cannibalising your highest-value terms first. Cannibalisation on a term that drives ten visits a month is not your priority. Cannibalisation on a term that should be driving your most qualified traffic is.

Set a review cadence before you close the audit. A full audit is typically an annual exercise for most sites. But a quarterly review of your top fifty pages by traffic and conversion is a much lighter process that keeps you from letting performance drift go unnoticed for twelve months. The teams I have seen do this well treat it as a standing agenda item rather than a project with a start and end date.

The Metrics That Actually Matter in a Content Audit

Traffic is the metric most people lead with, and it is one of the weakest primary signals in a content audit. A page can drive significant traffic from queries that have no commercial relevance to your business. A page can have low traffic but contribute meaningfully to assisted conversions across your funnel. Traffic tells you about reach. It does not tell you about value.

The metrics I weight most heavily in an audit are: assisted conversions or goal completions, engagement rate relative to comparable pages, keyword ranking trajectory over the last six months rather than a single snapshot, and referring domain count as a proxy for earned authority.

I have judged marketing effectiveness work for the Effie Awards, and one thing that becomes clear when you spend time evaluating campaigns against business outcomes is how often marketers confuse activity metrics with impact metrics. The same confusion shows up in content audits. A team that is proud of its publishing volume is usually a team that has not asked whether any of that volume is doing commercial work.

One metric worth adding if your analytics setup supports it: scroll depth on long-form content. A page with a high average time on page but a low scroll depth is probably being abandoned at the introduction. That is a different problem from a page where users are reading to the end and then leaving without converting.

Common Audit Mistakes and How to Avoid Them

The most common mistake is using a twelve-month traffic threshold as the only removal criterion. This leads teams to keep pages that have no strategic value because they happened to pick up a handful of visits from a long-tail query. A page receiving forty visits a year from people who are not your audience and who never convert is not a page worth keeping.

The second most common mistake is auditing content in isolation from the site’s link profile. Before you remove any page, check whether it has inbound links. If it does, you need a redirect plan. Removing a page with meaningful link equity without redirecting it is throwing away authority you have already earned.

The third mistake is treating the audit as a one-time exercise. Sites change, search behaviour changes, and business priorities change. A content audit is a recurring process, not a project you complete and then forget about.

The fourth mistake, and the one I see most often in larger organisations, is running the audit without clear ownership of the decisions. If the output requires sign-off from five different stakeholders before a single page can be removed, the audit will stall. Establish decision rights before you start, not after you have produced a 600-row spreadsheet that nobody has the authority to act on.

Early in my agency career, I inherited a client account where the previous team had completed a content audit six months earlier and produced an excellent piece of analysis. Nothing had been done with it. The client had loved the document but had no internal process for acting on recommendations that touched multiple teams. The audit had answered the right questions and then sat in a shared drive. That experience shaped how I approach audit deliverables: the document is not the output. The decisions and the sequenced action plan are the output.

If you are building a content programme from the ground up rather than auditing an existing one, the broader thinking on editorial planning and audience strategy in the Content Strategy & Editorial hub will give you the framework to build something that does not need a major remediation exercise in three years’ time.

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.

Frequently Asked Questions