What freshness means for AI answers

How it works

When an assistant answers a question with a web search, it picks a few sources out of many. Besides relevance, it matters how current a page is: a price list from three years ago is a poor source for a question asked today.

The main way crawlers recognise freshness is the XML sitemap, a file that lists every page of your website; more on it under XML sitemap. Each entry can carry a change date, the lastmod field:

<url>
  <loc>https://example.com/pricing</loc>
  <lastmod>2026-09-14</lastmod>
</url>

Google uses lastmod if it is consistently accurate. If the date changes on every request or without a real change, it is ignored.

bytereport looks for the sitemap named in robots.txt (the first Sitemap: line), otherwise at /sitemap.xml, and reads the most recent lastmod of all entries.

What matters

• A reachable sitemap. If bytereport finds none, it warns.
• lastmod in the entries. Without a change date a crawler can't tell what is new; bytereport warns.
• The most recent entry no older than 90 days. If the last change is older, bytereport warns. That doesn't mean every page has to be renewed this often; it only shows whether anything on the website is moving at all.
• Honest dates. Set lastmod only for a real change to the content, not on every build.
• Dates on the page. A visible date (“updated on …”) and dateModified in structured data also help with articles and guides.

How to set it up

Most content management systems and SEO plugins generate the sitemap themselves and enter lastmod when a page is saved; check in the settings that the sitemap is switched on. Add its address to robots.txt:

Sitemap: https://example.com/sitemap.xml

For custom-built websites, your agency generates the sitemap on publishing and uses the date of the last content change as lastmod. Most important of all: keep the pages that change up to date, such as prices, offers, opening hours and contacts.