An XML sitemap is a list. Google has said for years it does not guarantee indexing. What it does is remove guesswork: here are the URLs we consider canonical and public.
What belongs
- Canonical, indexable URLs only — the ones you would be happy to see as a blue link.
- A real
lastmodwhen the content changed. Fake yesterday’s date on everything and the signal becomes noise. - A
Sitemap:line in robots.txt pointing at the file. Then submit the same URL in Search Console.
What does not
- noindexed URLs, thank-you pages, account pages, cart and checkout.
- Redirects. List the destination, not the hop.
- Parameter copies (
?color=navy) when the canonical is the clean URL. - Images and videos unless you are deliberately running image/video sitemaps. Do not dump every media file into the page sitemap.
How it goes wrong in CMS land
Yoast and Rank Math will generate a sitemap that includes every custom post type you forgot to exclude. WooCommerce will add attribute archives. One extra tick in a plugin and you have published 8,000 near-duplicate URLs. Open /sitemap_index.xml, count the children, and delete the types that should never rank.
Check it
The sitemap checker looks for a file at the usual paths and in robots.txt. It will not parse 50,000 URLs — that is Search Console’s job — but it will tell you if the door is missing. Pair it with the full audit so you are not sitemapping a noindexed site.