The list you hand Google
XML Sitemaps: What Google Does With Yours, With Examples, Limits and Generators
Somewhere a vendor is still selling sitemap submission to hundreds of search engines. The file is free, your platform probably writes it already, and Google reads it as a suggestion. That does not make it useless. A sitemap is the one place you tell Google, in writing, which pages you think matter.
October 26, 2026 · 10 min read
The short answer
An XML sitemap is a file that lists the URLs you want search engines to find, with optional dates and image, video or language details. Google treats it as a hint: it helps discovery and recrawling but guarantees neither crawling nor indexing. One file holds up to 50,000 URLs or 50MB uncompressed. Submit it in Search Console and name it in robots.txt.
On this page
- What a sitemap earns you, and what it never will
- A sitemap file, tag by tag
- Three formats, and the numbers that cap each file
- lastmod is the one date Google checks your work on
- Index files: split the list so it tells you something
- Where your platform already keeps a sitemap
- Submitting it, and the ping that stopped answering
- Keeping the list honest after launch
What a sitemap earns you, and what it never will
Google's sitemaps overview defines a sitemap as a file that gives it information about the pages, videos and other files on a site and how they relate. In practice it does three jobs: it helps Google find URLs it might not reach through links, it tells Google when pages changed, and it carries extra detail about images, video, news and language versions. Bing goes further in its July 2025 guidance, calling sitemaps a foundational signal for URL coverage across its search and Copilot experiences, while adding that no tool can guarantee how content appears in AI answers.
A sitemap helps with
- New pages on a site that few other sites link to yet
- Pages buried several clicks deep on a large site
- Recrawling a page you changed, if the date you give is honest
- Images and videos Google might otherwise miss
- Language and regional versions, listed as alternates
A sitemap does not
- Force Google to crawl or index anything
- Raise a page's ranking by a single position
- Rescue a thin, duplicate or noindexed page
- Override a robots.txt block
- Remove a page from Google when you delete the entry
Who can live without one
Google names three kinds of site that benefit most: large ones, new ones with few external links, and those heavy with video, images or news. It also says a site may not need one if it is small (roughly 500 pages or fewer), every important page is linked internally, and it carries few media or news files. A twelve-page site for a pest control company, with every page in the menu, gains little. Most platforms generate a sitemap anyway, so the real question is rarely whether you have one. It is whether yours is telling the truth.
A sitemap file, tag by tag
A minimal XML sitemap is shorter than most people expect. It opens with an XML declaration, wraps everything in one urlset element that declares the sitemaps.org namespace, and gives each page a url entry holding its address in loc and, optionally, a lastmod date. That is the entire required vocabulary. Everything else is an extension or a tag Google skips.
| Required? | What Google does with it | |
|---|---|---|
| urlset (with the sitemaps.org 0.9 namespace) | Yes, once | Marks the file as a sitemap and wraps every entry |
| url | Yes, one per page | Holds a single page's entry |
| loc | Yes | The full, absolute, canonical address, with characters such as ampersands entity-escaped |
| lastmod | No | Helps schedule recrawls, but only when the dates prove accurate over time |
| changefreq | No | Ignored by Google and by Bing |
| priority | No | Ignored by Google and by Bing |
| image:image with image:loc | No | Points to images on the page, up to 1,000 per url entry |
| xhtml:link with hreflang | No | Lists alternate language or regional versions of the page |
Three house rules from Google's guide to building a sitemap: save the file as UTF-8, use absolute URLs, and list only the canonical version of each page. Place it at the root, because unless you submit it in Search Console, a sitemap covers only URLs in its own directory and below. Google also dropped the caption, title, geo location and license image tags from its documentation, so a plugin still writing them is wasting bytes. Language alternates are covered in our international SEO guide, and which version counts as canonical in our canonical tags guide.
Three formats, and the numbers that cap each file
| What it can carry | Where it fits | |
|---|---|---|
| XML | Every URL, plus image, video, news and language details | Almost every site |
| RSS 2.0 or Atom 1.0 feed | Usually only recent URLs; mRSS can describe video, but no images or news | A blog that already publishes a feed, alongside an XML sitemap |
| Plain text | One URL per line in a .txt file, and nothing else | A tiny site with no generator |
- A single sitemap holds 50,000 URLs or 50MB uncompressed, whichever comes first. Compressing it with gzip saves bandwidth, but the limit applies to the uncompressed size.
- A sitemap index file may list up to 50,000 sitemaps. It cannot list other index files.
- Search Console accepts up to 500 sitemap index files for each site.
- A Google News sitemap may carry up to 1,000 news entries and should list only articles published in the last two days.
- WordPress core caps each of its sitemaps at 2,000 URLs by default, a number a developer can change with a filter.
Nobody running a local business site will meet these ceilings. A store with 80,000 product variants, each on its own URL, will hit the first one, and the ecommerce SEO section deals with whether all those variant URLs should exist at all.
lastmod is the one date Google checks your work on
Google says it uses lastmod only when the value is consistently and verifiably accurate, for example when it matches the page's real last change. It counts a change to the main content, the structured data or the links as significant. A new copyright year in the footer does not count. Bing's guidance lines up: the date should reflect when the content changed, not when the sitemap file was generated, preferably as a full ISO 8601 timestamp with the time included.
Now the comedy. The Next.js documentation's own sitemap example sets lastModified to new Date(), which stamps every URL with the second the file was built. Paste that into production and you tell Google every page on the site changed today, and again tomorrow. Plenty of plugins and online generators do the equivalent. Google learns your dates are noise and stops reading them, which is exactly backward for the day you really do rewrite the pricing page.
Index files: split the list so it tells you something
A sitemap index looks like a sitemap of sitemaps. It uses a sitemapindex element instead of urlset, and each entry holds the loc of one child sitemap plus an optional lastmod for when that file changed. Google's large sitemap documentation requires the child files to live on the same site as the index and in the same directory or lower, unless you have set up cross-site submission in Search Console.
Here is the practical reason to split even a modest site. Search Console lets you filter the Page indexing report by sitemap, so separate files for service pages, location pages, blog posts and products turn one vague number into a diagnosis. Say your services sitemap is fully indexed and your locations sitemap is half indexed. The location pages are the problem, and they are probably thin or near-duplicates, which our guide to programmatic SEO explains. One undivided file would only have told you something, somewhere, is wrong.
Where your platform already keeps a sitemap
Google says the best way to make a sitemap is to let your website software generate it, and most software does. Before you buy or build anything, look for the file your platform already writes.
| Where to find it | What to watch | |
|---|---|---|
| WordPress (core, since 5.5) | /wp-sitemap.xml, an index of smaller sitemaps | SEO plugins such as Yoast SEO write their own; submit only the one that is live. Core sitemaps switch off when the site discourages search engines. |
| Shopify | /sitemap.xml at the root of the store's domain | Generated and updated automatically, with separate files for products, collections, blog posts and pages. Shopify says Google cannot read it while the store is in private mode. |
| Next.js (App Router) | A sitemap.xml or sitemap.ts file in the app folder | generateSitemaps splits a large catalog into numbered files. Replace the example's new Date() with real dates. |
| Hand-built or static HTML | Wherever you upload it | A desktop crawler can build one. Screaming Frog's free SEO Spider handles up to 500 URLs and leaves out redirects, errors, noindexed and canonicalized pages by default. |
Free online generators that crawl your site once produce a snapshot, and a snapshot goes stale the afternoon you add a page. Google itself links to a list of third-party generators that it describes as old and unmaintained. A sitemap that writes itself from your content is the only kind that stays honest without someone remembering to redo it.
Submitting it, and the ping that stopped answering
- 1
Open the file yourself
Load the sitemap URL in a browser. It should show XML, list your canonical protocol and hostname, and contain no addresses you would be embarrassed to rank.
- 2
Name it in robots.txt
Add a Sitemap line with the full URL. Google picks it up the next time it crawls robots.txt, and it works even for people without Search Console access.
- 3
Submit it in Search Console
In the Sitemaps report, paste the URL and submit. You need owner permission on the property. Remember that http, https, www and bare-domain variants are separate URL-prefix properties; our Search Console guide explains why a Domain property avoids that.
- 4
Give Bing the same file
Submit it in Bing Webmaster Tools. Bing pairs sitemaps, for full coverage, with IndexNow, which notifies participating engines the moment a URL is added, updated or deleted.
- 5
Read the status
Success, Has errors or Couldn't fetch. If it is anything but Success, our guide to XML sitemap errors walks through each failure.
- 6
Come back in a few weeks
Filter the Page indexing report by that sitemap and compare indexed pages with submitted ones. The gap is your to-do list.
Two old habits to drop. Google retired its sitemap ping endpoint, announced in June 2023; requests to it now return a 404, and a plugin still pinging does no harm and no good. And deleting a sitemap in Search Console only removes it from the report. Google does not forget the sitemap or its URLs; to stop crawling, block or remove the pages themselves. If a submitted file keeps failing to load, Google retries for a few days, then gives up.
Keeping the list honest after launch
A sitemap goes wrong quietly. A redesign renames URLs and the old ones linger, a plugin adds tag archives, someone noindexes a page and the generator keeps listing it. Each entry Google fetches that redirects or refuses indexing teaches it to trust the list a little less. Check the file monthly as part of routine SEO maintenance, and after any website migration, confirm that every listed URL returns a 200 and that redirects are not hiding inside it.
If the sitemap is clean and pages still are not indexed, the file was never the blocker; our fix for a site not indexed by Google and the technical SEO checklist cover the rest, and sites built in JavaScript should also read JavaScript SEO. Or hand it off. Our free website audit emails a PDF of SEO, speed, mobile, security and content fixes, and sitemap hygiene is routine technical work in our SEO programs. The SEO guide shows where it sits among everything else.
Questions owners ask about SEO
Not strictly. Google says a site of roughly 500 pages or fewer, with every important page linked internally and few media files, may not need one. It costs nothing, though, and most platforms write one automatically. The rule that matters is accuracy: a small, correct sitemap beats a large, stale one every time.
No. Google calls a sitemap a hint that helps it find and recrawl pages. It has no effect on where a page ranks once indexed, and it cannot make Google index a page it considers thin or duplicate. Rankings come from relevance, quality and links, which the rest of the SEO guide covers.
An XML sitemap is a machine-readable file for search engines, usually never seen by visitors. An HTML sitemap is an ordinary page of links for people, which crawlers can also follow. Large sites sometimes keep both. For most small businesses, a clear menu and footer do the HTML sitemap's job.
Leave out anything you do not want shown in search: redirected URLs, noindexed pages, non-canonical duplicates, filter and parameter URLs, thank-you pages, login screens and internal search results. List the one canonical address for each page that should rank. Our canonical tags guide helps decide which address that is.
Yes. You can submit several sitemaps and several index files, and Search Console accepts up to 500 index files per site. Splitting by page type makes the indexing report more useful. The mistake to avoid is two systems writing competing sitemaps, such as WordPress core and an SEO plugin, with one of them stale.
For Microsoft, yes. Bing's July 2025 guidance ties sitemaps and IndexNow to discovery for its AI experiences, including Copilot, while saying no tool guarantees an appearance in AI answers. Other AI companies have not published comparable guidance that we could verify. To see what the engines say about you today, run the AI visibility check.
Keep reading
- Google's free dashboardHow to Use Google Search Console: Setup, the Reports Worth Reading and a Monthly RoutineSet up Google Search Console the right way, read the reports that matter, check the new generative AI report and run five checks each month. Free and current.
- Under the hoodTechnical SEO: The Checklist That Decides Whether Google Can Read Your SiteTechnical SEO in plain English: robots.txt, sitemaps, canonicals, redirects, status codes, rendering and speed, plus a one-afternoon audit you can run free.
- The head of the documentMeta Tags for SEO: Which Ones Google Reads, and Which It IgnoresWhich HTML meta tags Google actually uses, which it ignores, and what each one does, with a copy-ready example of every tag worth keeping in your head.
- Inside the machineHow Search Engines Work: Crawling, Indexing and Ranking, Explained by Google's Own RulesHow Google and Bing find, store and rank pages, in plain English: crawling, rendering, indexing, serving, and where AI Overviews fit. Sourced from Google.
- SEO by the numbersSEO Statistics for 2026: Search Share, Click Rates, AI Overviews and What SEO CostsSEO statistics traced to the publisher: Google's search share, CTR by position, zero-click and AI Overview studies, page age, and what SEO costs in 2026.
- Video searchYouTube SEO: How YouTube Says It Ranks Videos, and What a Business Can Do About ItHow YouTube search and recommendations work, in YouTube's own words, plus titles, thumbnails, chapters and retention for a business channel.
- Everything in SEO GuideSee the section
Get a sitemap Google can trust
Send us your site address. We will check what your sitemap lists against what Google has indexed and reply with what to add, cut and fix.
Which of these sounds like you? (pick any)
Rather talk now? Call (210) 346-0848.