XML sitemap
An XML sitemap is a machine-readable file listing the pages on a site that you want search engines to find, usually at /sitemap.xml. It's a discovery aid: it helps crawlers locate pages, though it's no promise that any of them will rank.
In practice
Search engines find pages mainly by following links. A sitemap supplements that by handing over the list directly, which matters most for pages that are newly published, buried several clicks deep, or not linked from anywhere prominent. On a small, well-linked site it adds little. On a large catalogue, or immediately after a restructure, it's genuinely useful.
What it does not do is promise anything. Listing a page requests discovery; whether it gets indexed and how it ranks depend on the page itself. A sitemap is not a way to rescue thin content, and submitting one after a traffic drop is treating a symptom.
The sitemap should contain only canonical, indexable URLs that return a working page. That sounds obvious and is routinely violated: sitemaps ship containing redirected URLs, pages excluded from indexing, or addresses that 404. Each of those is a small contradiction between what you're asking a crawler to fetch and what you actually want, and a sitemap full of them gets trusted less.
It earns its keep during a website migration. A fresh sitemap of the new URL structure, submitted as soon as the new site is live, measurably speeds up how quickly search engines discover the new addresses and process the redirects pointing at them. It's one of the few levers that genuinely accelerates recovery.
One check worth doing occasionally: open your sitemap and compare its URL count to the number of pages you believe you have. A large gap in either direction is informative. Far more entries usually means duplicates or parameter variants are leaking in. Far fewer means pages are being excluded that you may not have intended to exclude.
Common questions
Do I need an XML sitemap?
Most sites benefit, and few are harmed. It matters most for large sites, newly launched pages, and anything recently restructured. For a small, well-linked site the effect is modest.
Does an XML sitemap improve rankings?
Not directly. It helps search engines discover pages. Whether those pages rank depends on the content, the site's authority, and competition. Treat it as plumbing, not as an optimization.
What should not be in a sitemap?
Redirected URLs, pages excluded from indexing, duplicates, and anything returning an error. A sitemap should list only canonical, indexable URLs that load correctly.