What is Image sitemap?
An ordinary XML sitemap lists pages. The image extension attaches the images that appear on each page: one `image:image` element per image with its address in `image:loc`, up to 1,000 images per page. Google accepts either a separate image sitemap or the tags in the existing sitemap; both are equally fine. The image addresses may sit on another domain, for instance with a content delivery network, as long as both domains are verified in Search Console.
Google describes it as a way to report images it might not otherwise find, such as those that only reach the page through JavaScript. For images that sit in the source code as an ordinary `img` element, it is an additional signal, not a requirement. And a sitemap does not lift a block: if `robots.txt` disallows fetching the image files, Google will not fetch them despite the sitemap.
Since 6 August 2022, the old fields for caption, title, licence and location have had no effect. An image sitemap therefore only tells Google where an image is, not what it shows. Google learns that from the page itself: from the text nearby, the caption, the alt text and the file name.
Why does Image sitemap matter?
An image sitemap is the simplest way to point Google to images that a script loads later, for instance in galleries or product carousels. It does not replace context on the page, though: since August 2022 it carries only the image address, up to 1,000 per page, and no description.
Image sitemap in practice
- 01A shop whose product galleries load via JavaScript adds the addresses of all gallery images to its sitemap for each product page.
- 02A website serves its images from a content delivery network on a separate domain and also verifies that domain in Search Console, as Google requires for image addresses on another domain.
- 03A sitemap plugin still writes image titles and captions into the sitemap. That does no harm, but Google has not evaluated these fields since 2022.
