Running a Content Audit Without Losing a Week

Last reviewed on 2026-05-12.

Content audits have a reputation problem. The phrase suggests something exhaustive and slow: every URL, every metric, a sprawling spreadsheet that takes a quarter to fill in and never gets acted on. That kind of audit is real, and it is usually a poor use of time. The version that actually moves the needle is much smaller.

An audit is useful when it answers a specific question: what should we do with the content we already have? Everything else is decoration. If the output of the audit is not a clear backlog of actions, the audit failed regardless of how thorough it looked.

Scope the audit before doing the audit

The first decision is what is in scope. For most teams, "all our content" is the wrong scope. A tighter scope produces a faster, more useful audit.

  • Top-traffic pages. The pages that drive 80% of organic traffic. Usually fewer than fifty.
  • Money pages. Pricing, comparison, key product pages — anything where bad copy directly costs you.
  • Topic cluster. All pages on one specific topic you are trying to own. Useful when you are about to invest in deeper coverage of that topic.

Doing an audit of all three at once is fine. Doing one of them well in a week is better than doing all of them at once over a month.

The data you actually need

Most audit templates ask for too much information. The fields below are usually enough to make decisions:

  • URL
  • Title and primary intent (informational, commercial, navigational)
  • Organic clicks and impressions over the last six months
  • Current ranking position for the page's primary query
  • Word count and last meaningful update date
  • Whether the page has internal links pointing to it (and from where)
  • One-line note on quality and accuracy as it stands today

Anything beyond that field set tends to slow the audit down without changing the conclusions. Add fields back only if you are sure they will change a decision.

A decision tree, not a scorecard

Scorecards (give each page a 1–10) feel objective and are mostly noise. A decision tree is faster and produces clearer outcomes. For each page, ask in order:

  1. Is this page bringing in traffic or conversions today? If yes, default to keep and improve. If no, continue.
  2. Is there a closely related page that already does this job? If yes, merge the two into the stronger URL and redirect.
  3. Is the topic still relevant to the business? If no, delete and either redirect to a parent page or return a 410.
  4. Could the page rank or convert with a real update? If yes, update: rewrite the lead, refresh the data, fix the structure, strengthen internal links.
  5. If none of the above is true, keep but deprioritise: leave it alone and check again next time.

That is the entire decision framework. Audits get stuck when teams try to make this tree more sophisticated than it needs to be.

The two parts that are easy to skip

Two things drop out of audits that should not:

Internal linking. When you delete or merge a page, the internal links pointing at it have to go somewhere. The pages doing the linking usually need light updates too. Skip this step and you accumulate dead-end links faster than you fix them.

Why the page underperformed. "Update" is a verb, not a plan. A page that did not rank usually has a real reason — wrong intent match, thin coverage relative to the SERP, weak topical relevance, no internal links. Naming the reason on the audit row makes the rewrite brief much easier later.

A worked example

Imagine a fifty-page B2B blog. After the audit, the breakdown might look like:

  • Six pages bringing in nearly all the traffic. Keep and improve over the next quarter — these are the ones to give the most attention to.
  • Twelve pages on closely related topics that should be merged into four stronger pillar pages.
  • Ten pages on topics no longer relevant to the current product positioning. Delete; redirect to the parent category.
  • Fifteen pages that could rank with a real update. Rewrite, scheduled across the next quarter.
  • Seven pages that are fine but inactive. Leave alone until the next audit.

That output is a backlog. It can go straight into a content calendar, prioritised by traffic potential or business importance.

Common mistakes

A few patterns that usually mean an audit will stall:

  • Auditing in isolation. Without sales, product, and SEO input, the audit produces decisions other teams overturn later.
  • No clear approver. An audit with five voices and no decider produces analysis, not action.
  • Over-quantifying. Every page rated on six axes from 1–10. None of those ratings get used.
  • Skipping deletes. Deleting pages feels harsh; leaving low-quality pages indexed harms the rest of the site. Default to delete-or-merge for the weakest tail.

How long should this take?

For fifty pages with a clean dataset, the audit itself is one to two days of focused work, plus a half-day of internal review. Building the backlog and prioritising adds another day. Two-and-a-half days of real work; not the quarter-long project the phrase "content audit" sometimes implies.

If an audit is taking much longer than that, the scope is probably wrong. Cut to the top-traffic pages, finish that, and let the broader audit come later — or never.

Where this fits

Audits are upstream of briefs: the backlog they produce is what you brief writers against. They are downstream of KPI choices: what you measure determines what counts as a successful page. For an applied example, see the SaaS content programme case study.

If you would like help running an audit on a specific topic cluster or section of your site, the next step is a short conversation. Email [email protected] or use the contact form.