Sitemaps are one of the oldest technical SEO artifacts and among the most over-credited. They help with a specific, narrow problem. They do not help with the problem most people deploy them against.
What a sitemap is
An XML file listing URLs you would like a search engine to know about, optionally with a last-modified date, a change frequency, and a priority.
Two of those four are effectively ignored. Google has stated it does not use <priority> and generally disregards <changefreq>. <lastmod> is used — but only when it is consistently accurate. Sites that bump lastmod on every page at every deploy teach Google that the field is meaningless on that site, and it gets discounted.
What it does
Aids discovery. For URLs not well linked internally, a sitemap gives Google a path to them. This is the real function.
Provides freshness hints. An accurate lastmod helps prioritize re-crawling of changed pages.
Enables reporting. Search Console reports coverage per submitted sitemap, which makes sitemaps a useful diagnostic unit — one per content type gives you a per-type indexing rate.
What it does not do
It does not cause indexing. Inclusion is not a request Google is obliged to honor, and it does not override a quality judgment. If pages show "Crawled – currently not indexed," resubmitting the sitemap changes nothing — Google found them, crawled them, and declined. The problem is the pages.
It does not fix crawl budget. Sitemaps do not increase how much Google crawls. They redirect attention within whatever budget exists.
It does not overcome bad architecture. A URL reachable only via the sitemap and linked from nowhere is a URL Google has been told about and given no reason to value. Sitemap inclusion is not a substitute for internal links; it is a complement to them.
That last point is where the most common misuse sits. Teams facing an indexing problem submit a sitemap, see no change, and conclude something mysterious is happening. What is happening is that discovery was never the constraint.
Getting it right
Only include canonical, indexable, 200-status URLs. A sitemap containing redirects, 404s, noindex pages, or non-canonical variants sends contradictory signals — you are asking Google to index pages you have told it not to index. Search Console flags these, and a high error rate reduces trust in the whole file.
Keep it current. Publish a page, add it. Retire a page, remove it. A stale sitemap listing removed URLs generates avoidable 404 crawls.
Respect the limits. 50,000 URLs or 50MB uncompressed per file. Beyond that, use a sitemap index pointing at multiple files.
Split by type. Separate sitemaps for products, categories, blog posts, and static pages cost nothing and give you per-type indexing rates in Search Console. When indexing degrades, that split tells you where immediately, which is usually most of the diagnosis.
Reference it from robots.txt. One line, discovered by every crawler that reads the file:
Sitemap: https://example.com/sitemap.xmlSubmit it in Search Console. Required to get the coverage reporting, which is the main operational benefit.
The diagnostic use
The most valuable thing a sitemap does is give you a denominator.
Submitted URLs versus indexed URLs, per sitemap, is a clean indexing rate. Track it over time and per content type. A healthy site sits high and stable. A drop in one type's rate is a specific, localized signal — something changed about that template, that section, or that content type — and you learn it from a report you get for free.
Without split sitemaps, the same degradation shows up as a vague site-wide number that tells you something is wrong without telling you where.
Specialized types worth knowing
Image and video sitemaps provide metadata that helps those items appear in the corresponding search surfaces. Worth it if images or video are a meaningful acquisition channel; otherwise not.
News sitemaps are required for Google News inclusion and have their own rules, including a 48-hour article window.
hreflang in sitemaps is the most maintainable way to handle international annotations at scale, as covered separately.
The honest summary
Generate it automatically, keep it accurate, split it by content type, reference it from robots.txt, submit it, and then use the coverage report as an ongoing diagnostic.
What you should not do is reach for it when pages are not being indexed. That symptom points at content quality, internal linking, or duplication — and a sitemap addresses none of the three.