Tools · check
Type a site or its sitemap’s address. The check reads robots.txt as Googlebot reads it, finds the sitemap, opens up to 5 of its files and reads 20 of the pages they list, then marks what keeps a listed page out of search: an error, a redirect, a noindex, a canonical that points elsewhere, or a robots.txt block.
Most platforms build the sitemap for you and put it at one of three addresses, which is why the check tries those after robots.txt. Here is where each one lives, and what each leaves out on purpose.
An index at the root of every domain the store uses, linking to separate sitemaps for products, collections, blogs and pages.
A store with international domains has sitemap files on each domain, and each is submitted on its own; the languages are added for you.
While the store is behind its password page, Google cannot read the sitemap.
An index WordPress has built since version 5.5, with at most 2,000 entries in each sitemap, a number a developer can change.
The robots.txt WordPress serves names the index.
Setting the site to discourage search engines turns the sitemaps off.
A request for /sitemap.xml redirects to the index, and each sitemap holds up to 1,000 entries unless a developer changes it.
Content types set to noindex are left out, which is what a sitemap should do: list only the pages you want in search.
Every sitemap file is sent with an X-Robots-Tag header of noindex, follow. That keeps the file itself out of search results and does nothing to the pages it lists.
An index of its own, 200 links in each sitemap unless you change it, with images included.
An index of separate sitemaps for the site’s pages, blog posts and store products, updated whenever the site changes.
Every page with its image details, usually updated within an hour of a change and at most a day, and not editable.
Disabled pages, password-protected pages and pages hidden from search are left out; trial sites and parking pages have no sitemap.
The report names a fault in a few words and leaves the finding to you. Here are eight of its lines, what each one means, and what the check above shows for it. It also marks Leading whitespace, Invalid URL, Incorrect namespace and Empty sitemap, in the report’s own words.
Couldn’t fetchGoogle could not get the file at all. The usual causes are an address that answers 404, a server that errors or times out, and a robots.txt rule that blocks the sitemap’s own address. The check reads the address the way Googlebot does and says which of the three it is. Submit the exact address it found.
URL not allowedThe sitemap lists addresses on another domain, or above its own folder: a sitemap at /blog/sitemap.xml may list only addresses under /blog/, unless it is submitted in Search Console or named in robots.txt. The check counts both. An address on another host is a fault; one above the folder is a note, because the check cannot see what you have submitted.
Path mismatchThe same fault in its most common form, reported as Missing www or Includes www: the sitemap sits on example.com and lists www.example.com, or the other way round. Pick the host the site redirects to, and use it for the sitemap and for every address in it.
Nested sitemap indexesAn index that lists another index. An index may name sitemaps, and a sitemap names pages; there is no third level. The check marks any file an index names that turns out to be an index itself.
Sitemap contains urls which are blocked by robots.txtThe sitemap asks Google to crawl pages that robots.txt tells it not to crawl. Either those pages do not belong in the sitemap, or the Disallow rule reaches further than it was meant to. The check runs up to 2,000 addresses from each file through the rules that apply to Googlebot and names the rule that blocks the first one.
Invalid dateA lastmod that is not a W3C date, such as 2026-09-15 or 2026-09-15T10:30:00+00:00. Google uses lastmod only when it is consistently and verifiably accurate, so a build date stamped on every page tells it nothing, even when it parses. The check says so when every address in a file carries the same date.
Parsing errorUsually an ampersand in an address written as a bare & instead of &. One is enough to make the whole file invalid XML, and the check counts them.
Too many URLsMore than 50,000 addresses in one file. Sitemap file size error is the same limit by weight: 50 MB before compression. Split the list and name the parts in an index. The check counts the entries and reads the file’s size from its headers.
Open a sitemap in a browser and you often get a tidy table instead of code. The table comes from a stylesheet the file asks for in its first lines, <?xml-stylesheet type="text/xsl" href="…"?>, which the browser runs as XSLT. WordPress’s own sitemaps have come with one since version 5.5, and Yoast SEO’s sitemaps point to one too. The check notes the line on every file that carries it.
For search, nothing changes. Google reads the XML, not the table. Joost de Valk, who founded Yoast, puts it plainly: “Your XML sitemap still helps search engines find your pages.”
For people, the table goes. Anyone who opens the sitemap in Chrome sees the file’s code instead of a list. If clients or colleagues use the sitemap as an index of the site’s pages, give them an ordinary page of links instead. Mozilla and Apple have both backed removing XSLT from Firefox and Safari as well.
Open /robots.txt on the site and look for a line that starts with Sitemap:, because most sites name theirs there. If there is none, try /sitemap.xml, then /sitemap_index.xml, which Yoast SEO and Rank Math use, then /wp-sitemap.xml, which WordPress builds on its own. The check tries them in that order and says where it found the sitemap. On a site you own, Search Console’s Sitemaps report lists the sitemaps you have submitted.
No. Google’s documentation says a site of about 500 pages or fewer, with its pages linked from one another, may not need one, and that a sitemap does not guarantee a page is crawled or indexed. It earns its place on a large site, a new site with few links pointing to it, and a site with a lot of video, images or news. Every platform listed above builds one anyway, so the useful question is whether yours is right.
Not necessarily. Google’s introduction to robots.txt says the address of a blocked page can still appear in search results, without a description, when other pages link to it. robots.txt stops crawling, not indexing. To keep a page out, let Google crawl it and give it a noindex, in a robots meta tag or an X-Robots-Tag header; a noindex on a page Google may not crawl is never seen.
The check reads the answers that rule a page out: an error, a redirect, a noindex, a canonical that names another address, a robots.txt block. If a page passes all five and is still missing, the decision is Google’s. Google says a sitemap does not guarantee indexing, and when it picks which of two addresses to index it counts a sitemap as a weak signal, below redirects and canonical tags. Search Console’s URL Inspection tool shows the reason it gives for a single page.
Because a site keeps them out of search on purpose, and the check knows them: the cart, the checkout and the thank-you page after it, account pages and the pages under them, sign-in, sign-up, password and sign-out pages, wishlists, order tracking, site search, admin and preview addresses, email preferences, feeds and embeds. It knows them from the classes WooCommerce, Easy Digital Downloads, Magento, WordPress and Shopify themes put on the page, and from the addresses those platforms, BigCommerce and PrestaShop give them. Kept out with a noindex, a redirect, a canonical or a refusal, such a page is named and not counted. A WooCommerce checkout is read again with one product in a throwaway cart, the way the checkout speed test does, and nothing is ordered. A missing page, a server error or no answer still counts, whatever the page is.
Not to Google. Its sitemap documentation says Google ignores <priority> and <changefreq>. The protocol still allows both and other search engines may read them, so the check does not count them as faults. lastmod is the tag Google does use, and only when the dates are true.
There is nothing left to ping. Google announced in June 2023 that its sitemaps ping endpoint would stop working six months later, and a ping now gets a 404. Submit the sitemap once in Search Console or name it in robots.txt, and keep each lastmod true to the day the page last changed, which is what Google recommends instead.
A firewall or a bot protection service is refusing a crawler it does not know. The check names itself in its user agent, NorvahlCheck, with a link to this page, and some services answer that with 403 or 429 while they let Googlebot through. When every request is refused, the check says so instead of calling the sitemap missing. Search Console’s URL Inspection shows what Googlebot itself got; if Googlebot is refused too, that is the fault to fix first.
So that a check never weighs on your server. It reads up to 5 files and 20 of the pages they list, spread across each file, five pages at a time, for their status and their head, without following redirects. The robots.txt rules are checked against up to 2,000 addresses in every file, which costs your site nothing because that part runs in your browser. For every page on a large site, Search Console’s Page indexing report is the full list.
The address checked, what robots.txt and each file answered, the addresses of the files and pages it read with their status codes, and a salted hash of your connection that counts toward the limits: 15 checks a day, and a site at most 6 times in ten minutes. It keeps no file and no page. The files come to your browser to be judged and are gone when you close the tab. On a WooCommerce store it may put one product in a throwaway cart to read the checkout; that cart is the store’s own and expires with its session.
Next step · Norvahl, one designer-engineer, both tradesTaking on projects
Send me the address and what Search Console’s Sitemaps report says about it. I read both myself and reply within a working day.
reply within a working day · me, not an assistant · a 30 minute call, free · no newsletter · hello@norvahl.com