Skip to content
← Insights
Nov 14, 20265 min read

Keep your feed and sitemap on the same article version

Preserve article identity across edits and derive feed content, sitemap dates and structured metadata from the same published record.

An article folder connected to a feed tray and a small map card beside an unnumbered date stamp.

An article can display one title while its feed keeps an older version. A modification date that changes with every deployment adds another inconsistent signal. Each representation may parse successfully while giving subscribers a different account of what changed. Machine-readable publishing needs a shared content record behind those outputs.

For a small publication, begin with one article and inspect what each output says about it. The goal is consistency that a reader can rely on. A feed gives a subscriber access to updates. A sitemap supplies URLs to a search service. Article metadata describes the page. Each has a defined role in publishing.

Keep the article identity stable

Give the article a stable public URL and retain a source identity inside the publishing system. Editing a title should leave the same article attached to its history. Otherwise, an ordinary correction can accidentally create a second entry that the publication later has to reconcile.

The RSS 2.0 specification defines an item's GUID as a string that uniquely identifies that item. Feed readers can use it to decide whether they have already received the item. A publishing system should choose that identity deliberately and preserve it across routine edits.

A migration deserves particular attention. Compare an existing item's identifier before and after moving the content between systems. If the article keeps its identity in the CMS but receives a new GUID in every feed build, a subscriber may see repeated entries. Retaining a known identifier gives you a concrete behaviour to verify.

Derive the feed from published content

A feed should draw from the content state intended for the public. Check how your publishing system distinguishes a saved draft from an approved publication. A draft can exist in the CMS while remaining outside the public feed. Test the boundary using the system's documented content states.

Choose whether each feed item contains the complete article or a summary. Make that choice predictable for subscribers. If the feed contains a summary, ensure the link reaches the full public article. A summary that describes an earlier conclusion needs an update when the maintained article changes substantially.

Inspect the generated XML as well as its content. Characters in a title or link can require escaping. A single malformed item can affect a reader's ability to parse the feed. Use the feed format's validation tools, then open the feed in a reader to inspect how an ordinary subscription behaves.

Give the sitemap accurate dates

Google's sitemap construction guidance explains the role of the last-modified value. It should represent a significant change to the page when used. A website build can occur because an unrelated page changed, so the build timestamp needs a different place in your deployment records.

For an article, keep the date of the editorial change in the content record. Derive the sitemap entry from that value when it accurately represents the page update. If a deployment rebuilds the whole site, verify that unchanged articles retain their earlier modification dates. That keeps the signal attached to actual content history.

Include the public URL you want a reader to reach. Check for draft routes or obsolete addresses in the output. A sitemap entry that leads through an unintended redirect can reveal a mismatch between the route generator and the content model. Fix the shared rule so the next build produces consistent addresses.

Align article metadata with the visible page

Google's Article structured-data documentation describes fields for the article's identity and dates. Populate those fields from the same published record that supplies the visible heading. This reduces the chance that a title correction reaches the page while leaving a contradictory machine-readable headline.

Make the author reference specific to the person or organisation responsible for the article. Check any linked author page publicly. An empty profile route can remain unnoticed when the main article renders normally. Metadata verification needs to follow the referenced URL as well as inspect the markup itself.

Structured data can help a service understand a page under its documented rules. A valid record leaves search appearance and eventual citations to the receiving service's own process. Keep the acceptance check focused on the output you control: accurate information that matches the article a visitor can read.

A withdrawn article needs the same shared-state review. Determine what the publication intends to show at its old address and remove obsolete references from generated outputs where appropriate. Check the feed's behaviour for existing subscribers before assuming a removal erases their cached copy. Document the public result the publisher intends to maintain, then verify that result after the change is deployed.

Test publication as a state change

After a new publication, request the public article and its feed entry. Check the sitemap URL and any article metadata against the same expected record. A successful deployment job establishes that the build ran. The public responses establish which version readers and external services can actually receive.

Repeat the check after a substantive correction. Confirm that the article keeps its identity while the appropriate modification record changes. If the feed uses full text, inspect the corrected passage there too. A site with several representations benefits from checking all of them at the moment an edit becomes public.

Keep a small verification receipt with the article identity and the public URLs tested. It makes later complaints easier to investigate without duplicating the CMS as a second source of truth. Publishing becomes easier to maintain when every output describes the same article and every update has an observable public result.

PublishingRSSWeb Development