A content inventory assembled from sitemap and CMS data
The sitemap lists what the site wants crawlers to find. The CMS lists what editors have made. Neither is the real inventory. Put them side by side, though, and things start falling out: pages that exist but aren’t linked, entries missing from the sitemap, old URLs that still matter, fields that don’t match the live page.
The goal isn’t the biggest spreadsheet in the world. It’s a reliable view of the pages that matter for one decision: a migration, content ownership, an indexation review or a cleanup. Start with that and the inventory stays useful.
Choose the sources deliberately
For a basic inventory, collect the current sitemap, a CMS export and a crawl of the public site. Each source has a different job.
The sitemap shows URLs declared for discovery. It may be split across several files or include odd paths. The CMS export can show titles, status, content type, owner and last-edited fields, though those can be empty or misleading. A crawl shows what a crawler can reach from its starting point, and the status and canonical signals it sees.
Add Search Console or analytics data only when the question needs performance context. Those datasets have different coverage and definitions, so let them enrich the inventory rather than quietly become its source of truth. Keep the date range and export date with them.
Before combining files, check each source covers the same site, market and time period. A staging sitemap or an old CMS backup makes the inventory look inconsistent when the real problem is that the inputs describe different versions of the site.
Create a stable page key
Normalise the URL carefully. Standardise host and protocol where appropriate, remove tracking parameters only when they don’t identify a meaningful page, and handle trailing slashes consistently. I always keep the raw URL in its own column so I can trace every transformation.
Don’t collapse URLs just because they look alike. Case, parameters, language paths and product variants can change how a system treats a route. Decide how to handle each based on the site setup and the inventory’s purpose. If the same content sits under several URLs, keep those rows until you understand why.
A stable key lets you join a sitemap URL to a CMS entry and a crawl result. Sometimes the match is exact; sometimes you need a redirect map or a content ID. Be wary of fuzzy matching as an automatic fix. It can connect the wrong page and make the cleanup harder.
Keep source facts separate from interpretations
Build one tab or dataset with the raw source fields and another with derived decisions. Useful raw columns can include source name, URL, CMS identifier, page type, status, canonical URL, sitemap inclusion and crawl date. Add business owner or review date only if someone can maintain those fields.
Derived fields might flag missing matches, duplicate routes, stale content or pages needing a manual decision. Call them flags, not facts. “No CMS match found” is accurate. “Orphan page” is a conclusion that needs a crawl, an internal link review and a check for pages managed outside the CMS.
This distinction prevents a spreadsheet from turning assumptions into a work queue. A page absent from a CMS export may be generated dynamically or managed elsewhere. A URL missing from a sitemap may be intentionally excluded, or it may point to a forgotten gap. Investigate before deleting or redirecting it.
Check the joins before trusting the counts
After combining sources, reconcile the totals. Count unique URLs in each input, duplicate keys, unmatched rows and rows with several candidate matches. Inspect samples from every category, especially accented characters, encoded paths, query strings and language prefixes. I’d rather find a join error now than after the cleanup.
Check that redirects are represented clearly. A CMS entry may publish at a new route while the sitemap still contains an old one. A crawl may follow the redirect and report only the destination. Keep both the requested URL and final URL where possible so the relationship is visible.
If a URL appears in several sitemaps, that isn’t several pages. And don’t add up sitemap totals and call it the site’s page count. One page can appear in multiple files, and not every listed URL is necessarily indexable.
Make the inventory answer a real question
A migration inventory should identify content to preserve, update, merge, redirect or retire, with a person responsible for the decision. A technical review might focus on pages missing from sitemaps, canonical inconsistencies or status patterns. A content audit may include editorial ownership and review needs.
Do not ask one inventory to serve every team at once. If marketing needs a content calendar and engineering needs a migration map, create shared source data and separate views. Use names that make sense to the people who will act on them; a cryptic code may be efficient for a query and useless to an editor.
Include the update date, source files and transformation notes. If another person refreshes the inventory next month, they should be able to tell which fields came from the CMS and which were added during analysis. Keep access and retention aligned with the organisation’s policies, especially if the export contains contributor details.
Turn the file into a maintained process
An inventory gets stale as soon as pages change. Decide when to refresh it: before a migration, during a recurring audit or after a major catalogue release. Do not promise that every row will stay current if no one owns it.
Related reading
- Building a content inventory from a sitemap and CMS export
- A GA4 key events audit after a website redesign
- How to verify key event data after a GA4 change
Start with one task, such as finding public pages missing from both the sitemap and a CMS export. Gather the sources, match them openly and inspect the unmatched examples. Which decision should the inventory make easier, and who owns it once the spreadsheet is shared?