Running & Maintaining a Site

Content Audit

Definition

A content audit is an inventory of a site's published material combined with a judgement about each item. The inventory half is mechanical: every URL, its title, file type and size, the date it last changed, and its traffic over a chosen period. The judgement half is not: for each row someone decides to keep, update, merge, redirect or remove. Audits live in a spreadsheet because the useful operation is sorting — by last-modified date to find stale material, by traffic to find pages nobody reads, by file size to find the assets that dominate transfer. The starting list normally comes from sitemap.xml plus a fresh crawl, since the two disagree in informative ways: anything in the crawl but not the sitemap is undeclared, and anything in the sitemap but not the crawl is unreachable.

Why It Matters

Sites accumulate. A three-year-old site typically carries pages describing prices that changed, staff who left and products that were withdrawn, and none of it announces itself. The cost is not only embarrassment: out-of-date pages compete with current ones in search results, and a visitor landing on last year's terms has no way of knowing. An audit is also the cheapest way to find every broken link, oversized asset and duplicated page in one pass, rather than one complaint at a time.

How It Works

Crawl the site for the reachable set, pull the declared set from sitemap.xml, and reconcile the two. Add analytics for the same period so every row carries a traffic figure, then add file size, which is where you find the 30 MB PDF quietly costing more than everything else combined. Cross-reference internal links to spot any orphan page — reachable by URL but linked from nowhere — since those either need a place in the navigation or need retiring. Then work down the list recording a decision per row, with a date and an owner, so the next audit starts from what was agreed rather than from scratch. Rows marked for removal need a redirect or a deliberate 410 rather than silent deletion, and rows marked for update need a review date attached or they simply come back next time.

Real-World Example

A training provider has 140 course handouts published under tutor-notes.99helpers.site. Sorted by last-modified date, 51 of them have not been touched in over two years; sorted by views, 30 have fewer than ten views in twelve months. Sixteen describe a syllabus that no longer runs and are redirected to the current course index. The rest are left alone with a review date attached, and the whole exercise takes two people a day.

Common Mistakes

  • Auditing pages and ignoring files — the PDFs, images and downloads are usually the largest, oldest and least reviewed part of a site
  • Producing the inventory and stopping there — a spreadsheet with no decision column is a report, not an audit, and nothing changes as a result
  • Deleting low-traffic pages on the view count alone — a page with 40 views a year can be the one that closes a deal, and views do not measure that

Related Terms

Put a file online in seconds

Drop in a document, an image, a page or a whole static website and share the link — free, with no build step and no server to set up.

Host a file free →