Mueller can't confirm splitting sitemaps by age changes crawling

Google’s John Mueller listed five reasons sites split their XML sitemaps into several files, and he marked one of them as unproven, according to Search Engine Journal’s report. The unproven one is splitting by content age. Evergreen pages go in their own file, on the theory that a search engine can check that file less often. Mueller said a search engine theoretically might not need to revisit the older file as often, then said he does not know whether that actually happens.

He was answering an SEO who asked why anyone would trade one file for several. Mueller’s five reasons:

• Reporting by page type. A site keeps product detail pages and category pages in separate files to track each group. Mueller noted that Search Console’s Page indexing report can partly do this already.
• Freshness. Evergreen content sits in its own file so it might be checked less often. Mueller could not confirm that any search engine acts on the split.
• Splitting early. A site splits before it reaches 50,000 URLs so it never has to rework its setup in a hurry.
• hreflang. Sitemaps that carry hreflang annotations take so much space that 50,000 URLs can make a file too large.
• Nobody decided. Software generated several files and no one knows why.

The file limits are the reason that holds up

Two of the five reasons come straight from the sitemaps.org protocol, which caps each sitemap file at 50,000 URLs and 50MB. A site past either limit has to move to a sitemap index, a file that lists other sitemap files. An index can itself list up to 50,000 sitemaps. The likely gain from splitting early is that the site submits one index URL once and adds files behind it as it grows, instead of restructuring under pressure.

Sites with hreflang in their sitemaps hit the size limit first. Each URL entry lists every language version of the page, so entries grow with the number of languages. Take a made-up example: a site in 25 languages writes 25 alternate links under every URL. Its file reaches the byte limit long before the URL limit. For those sites the file size, not the URL count, decides when to split.

Why the freshness split is unproven

The protocol does describe a way the freshness split could work. Each sitemap listed in an index can carry an optional lastmod date, and the protocol says the date lets crawlers retrieve only the sitemaps modified since a certain date. A file of evergreen pages that rarely changes would keep an old date and could be skipped.

The protocol allows crawlers to do that. It does not say Google does, and Mueller said he does not know. Until Google says otherwise, treat a freshness split as a way to organize files, not as a way to change crawl rates.

The split also has a chance of working only if the index dates are accurate. A generator that stamps every sitemap with the build date makes the evergreen file look new on every build. The protocol’s rule for individual URLs points the same way: lastmod is the date the page last changed, not the date the sitemap was generated.

What to do

• Keep one file if the site is small. Search Engine Journal’s piece makes the same point.
• Split into an index before a file nears 50,000 URLs or 50MB. If the sitemaps carry hreflang, watch the file size, not only the URL count.
• Split by page type if you want indexing numbers per group, then compare the groups in the Page indexing report.
• Do not split by content age expecting Google to crawl the old file less or the new one more. If you already have that split, set each lastmod only when the file’s URLs actually change.
• If a plugin or CMS created the split, open the index and confirm every listed file still exists and holds live URLs.

An index file split by page type looks like this, with each lastmod set to the date that file’s URLs last changed:

<?xml version="1.0" encoding="UTF-8"?>
<sitemapindex xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
  <sitemap>
    <loc>https://www.example.com/sitemap-products.xml</loc>
    <lastmod>2026-03-28</lastmod>
  </sitemap>
  <sitemap>
    <loc>https://www.example.com/sitemap-categories.xml</loc>
    <lastmod>2026-03-02</lastmod>
  </sitemap>
</sitemapindex>