The Content Freshness Checker finds the dates a crawler can read on a page: article published and modified times, dates in structured data, the Last-Modified header, and any year near the top of the text. It then works out roughly how many days old the page is, with a pass up to 365 days, a warning up to 730 and a fail beyond that.
Updated
The headline says about how many days ago the page was last updated, or that it has no date a machine can read. Below it are six checks, one for each place a date can live and one for the overall age.
The age comes from the first date the tool finds, in this order: the modified time, the published time, the date in structured data, then the Last-Modified header.
| Source | What the tool reads |
|---|---|
| article:modified_time | A meta tag in the page head |
| article:published_time | A meta tag in the page head |
| Structured data | dateModified or datePublished in the page's JSON-LD |
| Last-Modified header | A header your server sends with the page |
| Visible year | A year from 2010 to 2029 in the first 4,000 characters of text |
Many sites send no Last-Modified header, which is common for pages built on the fly, so a warning on that line is not unusual. The meta tags and structured data are easier to add.
In Peak Answer's look at pages ChatGPT cites, only about 5% showed a visible date. A date helps readers judge a page, and a good result here does not mean more citations.
Peak Answer runs checks like these every week, then writes the content and gets the placements.