This translation has not been editorially reviewed yet. The German version is authoritative. Deutsche Fassung

GEO / AI

Why do ChatGPT and Google prefer web pages with a date?

ChatGPT, Google AI Overviews, and other answer engines favour web pages with a recognisable date, because they want to give their users current and reliable information. A page with no date could be five years old without anyone noticing. Our check therefore looks for a machine-readable publication or modification date in your page's code.

Freshness is a selection criterion for answer engines

When an answer engine chooses between two similarly good sources, it tends to pick the one that's visibly current. That makes sense: prices change, laws change, offers change. An answer from an undated source could be outdated, and that reflects back on the machine. A date works in two ways here. The publication date says when the content was created. The modification date says that someone takes care of it and maintains the text. Together, both paint the picture of a living, well-tended website — exactly the kind of source machines and people prefer. A visible addition like "Last updated: March 2026" further strengthens this picture for the human reader.

How the date becomes machine-readable

A visible date in the text isn't always enough for machines, since they can't reliably tell whether a year is the page date, a founding year, or a mention of an appointment. It becomes unambiguous via two technical routes our check recognises: dedicated meta fields in the page head for publication and modification time, or the "datePublished" and "dateModified" fields in the structured data, the page's invisible data sheet. Many content management systems maintain these fields automatically as soon as an article is saved. If that's not the case on your website, your agency can add the output with little effort. Either of the two routes is enough for full marks.

Ask your agency whether your pages output a machine-readable publication and modification date.
Turn on automatic maintenance of these date fields in your content management system.
On guide pages, also visibly show a line like "Last updated: Month Year".
Update content when prices, processes, or the legal situation change, instead of creating new pages.
Check after major revisions whether the modification date was updated along with the content.

Deutsche Fassung dieser Seite

FAQ

How does Google know how old a web page is?

Most reliably, from machine-readable date fields in the page code: meta tags in the page head, or the "datePublished" and "dateModified" fields in the structured data. Search engines also observe when a page's content actually changes. A visible date in the text helps readers, but doesn't replace the machine-readable attribution.

Should I just refresh the date on old pages?

No. A new date with no content revision is an empty claim, and machines recognise unchanged content from the text itself. Instead, actually update: check prices, processes, and examples, add anything new, cut anything outdated. Then the renewed modification date carries a real statement and strengthens trust in your page.

Does every page need a date, even the homepage?

Our check assesses every page measured individually, so a machine-readable date on every page is the goal. It's especially important for articles and guides, since there freshness decides citability. Content management systems usually maintain the field automatically on every save, so the homepage and service pages get a date too with no extra effort.