Sitemap Strategy for WooCommerce: A Curated List, Not an Export
What belongs in the sitemap, why lastmod becomes noise on synced stores, and how comparing sitemap count to indexed count diagnoses most indexation problems.
A sitemap is not a submission list you keep adding to. It is a statement about which URLs you consider worth indexing, and a store that puts everything in it is telling search engines that it has no opinion.
What belongs in a WooCommerce sitemap
Include:
- Products that are published, in stock or not, and indexable
- Categories and subcategories that contain products
- Blog posts and pages you want indexed
- Indexable facet pages, if you have deliberately chosen them
- Translated versions, each in its own language sitemap
Exclude:
- Cart, checkout, my-account, login
- Internal search results
- Sorting and filter parameters that are not indexable
- Tag archives, unless curated
- Attachment pages, author archives, date archives
- Empty categories
- noindex pages — including them is a contradictory signal
That last item is the most common error: a plugin adds every product to the sitemap, then a rule noindexes out-of-stock ones. The sitemap says “index this”, the page says “do not”. Crawlers resolve the contradiction on the safe side and crawl less.
Structure: one sitemap per language, not one for everything
On a multilingual store, a sitemap index pointing at per-language sitemaps is better than one combined file. It makes indexation measurable per market: you can see that German products are indexed and Polish ones are not, which is otherwise invisible.
Within each language, split by type: products, categories, content. A store with 20,000 products benefits from product sitemaps split at 5,000 URLs each, so one error does not invalidate a large file.
lastmod: the field everyone gets wrong
lastmod should change when the page meaningfully changes. In practice, plugins update it on every save — including saves triggered by a stock sync or a bulk price update, which touch every product daily.
The consequence: lastmod becomes a constant, search engines learn to ignore it, and you lose the ability to signal real updates. On a store running frequent sync jobs, this happens automatically without anyone deciding it.
Two options:
- Exclude stock-only changes from
lastmod, if your plugin allows it - Accept that
lastmodis unreliable and stop relying on it
Either is fine. What does not work is assuming it is accurate when a sync runs hourly.
Stale sitemaps from old platforms
If the store was migrated, an old sitemap file often survives at a predictable path. Search engines keep fetching it, discover dead URLs, and spend crawl time on 404s.
Check for common leftovers: sitemap.xml, sitemap_index.xml, sitemap-products.xml, and any file name the previous platform used. Remove or redirect them, and confirm the one being submitted in Search Console is the current one.
What the sitemap tells you about the store
Comparing sitemap count to indexed count is one of the most useful diagnostics available, and it takes a minute:
| Pattern | What it suggests |
|---|---|
| Sitemap 3,000 / indexed 2,800 | Healthy |
| Sitemap 3,000 / indexed 800 | Quality problem or crawl constraint |
| Sitemap 40,000 / indexed 3,000 | Index bloat — sitemap includes URLs that should not be there |
| Sitemap 500 / indexed 900 | Pages indexed that are not in the sitemap, often the ones you did not intend |
The last row is worth attention: it usually means parameter URLs or archived pages are being indexed while your curated pages are not. That is covered in the crawl budget notes.
Submission discipline
Submit the index file once, and let it update. Resubmitting the same sitemap does not accelerate anything and gives no new information.
When you add a section — say, a new language — submit that specific sitemap so it is picked up without waiting for the next full crawl cycle.
Do not submit sitemaps for staging domains, and do not leave a staging sitemap publicly reachable. It is a common source of duplicate-content confusion after a launch.
The minimum viable policy
- One sitemap index, per language, split by content type
- Only indexable, canonical, 200-status URLs
- No parameters, no internal search, no empty categories
- Old platform sitemaps removed
- Monthly comparison of sitemap count against indexed count
None of this is difficult, but it requires someone to decide that the sitemap is a curated list rather than an export. Most stores never make that decision, which is why the file grows to 40,000 entries while 3,000 pages do the work.
If you inherited a store with a large, unmanaged sitemap, the cleanup is a few hours of configuration — and it makes every subsequent indexation question easier to answer. It is the first thing we check in a technical SEO audit, because it is the cheapest way to see what a site believes about itself.