Sitemap submission to Baidu

A sitemap is the standing declaration of what a site contains, and in Baidu it is the appropriate channel for the existing body of pages — as distinct from the push channels, which are for what is new.

How Baidu processes it

The file is fetched and processed cyclically, which makes it reliable and slower than push. That is the trade the platform’s own comparison sets out: no development needed, complete coverage, but not the route for something that must be indexed today.

What belongs in it

• Canonical URLs only — one form per page, matching what internal links use.
• Pages that should actually rank. A sitemap listing filter combinations and internal search URLs invites crawling of pages Baidu’s own rules grade as worthless.
• Accurate modification dates, so the crawler can prioritise what changed.

ETag and conditional fetching

Baidu documents the use of ETag to support sitemap submission — the server tells the crawler whether the file has changed since it was last fetched. On a large or frequently regenerated sitemap this avoids repeated downloads of an unchanged file, which is capacity returned to crawling actual pages.

The common failures

• The file unreachable, redirecting, or answering something other than 200.
• URLs that do not match the canonical form — parameter variants, protocol or www mismatches.
• A sitemap listing URLs blocked in robots.txt: an instruction to crawl and an instruction not to, in the same site.
• A stale file regenerated less often than the site publishes.

The arrangement that works

Regenerate on a schedule that matches publishing, keep it to canonical URLs, serve it with correct caching headers, and use the push channels for anything time-sensitive. The sitemap is the map; it is not the messenger.

Verification

Check in the platform that the file was fetched and processed, and compare the URL count against index volume. A large gap between submitted and indexed is a quality or duplication finding, not a sitemap problem.