toolkit

Scenario · Content and search

This copy feels old

“This copy feels old.” Or: “When did we last touch this page?”

"This copy feels old" is a hunch until somebody puts a date on it. This reads every real signal a page gives off about its own age — a stated publish date, a visible "last updated" line, the server's own timestamp — and builds an oldest-first list, grading a blog post and a legal page against very different expectations for how fast each should age. Old and accurate is still a perfectly good page; deciding whether a specific page has drifted out of date is a judgment only the person who owns the content can make.

What to ask for

Content freshnessContent-freshness audit — works out how old a page's content is from every freshness signal it can see: schema datePublished/dateModified, time elements, the Last-Modified header, and visible "last updated" dates.

See it work

A real run of Content freshness:

REVIEW — <date> · 5.2yr old
  [warn] freshest date is <date> — 1,<age> old, past the 730-day threshold for "default" content; schedule a refresh

  report: ./out/content-freshness-content-freshness.html
  json:   ./out/content-freshness.json

The captured report, exactly as a run hands it to a client —open the full report ↗

“Feels old” is a hunch until it’s a date. The freshness check reads every signal a page actually declares — schema datePublished/dateModified, a visible “last updated” line, the server’s own Last-Modified header — and trusts the content’s own claim over the server’s, since a template can report a fresh header on copy nobody has touched in years.

  1. Run the freshness check for the page or the full site’s URL list.
  2. Start with the oldest pages, and read the specific evidence behind each date — which signal it came from, and how old the check considers “too old” for that content type.
  3. [manual] Ask the content owner whether the page is still accurate and still belongs in the current service set — an old, correct page is not a defect, and this is the judgement no dated signal can make for them.
  4. [manual] For confirmed-stale pages, create the rewrite or review task, then rerun the check after publishing to record the new date and confirm it actually landed.

Old and accurate is a pass

The check never fails a page for age by itself. Its job is to build the oldest-first queue; whether each entry in it needs work is a judgement about accuracy, not a number the tool can supply on its own.

What this does not cover

Whether the copy is *wrong*. It reports how old each page's content actually is, oldest first, and never hard-fails a page for age alone — old and accurate is a perfectly good page. A legal or evergreen page ages slowly by design; a blog post or a promo ages fast, and the check grades each against a different threshold for exactly that reason.

← All scenarios