WooCommerce 11.0 carries more real breaking changes than the version number suggests, and the headline one is a removal rather than a deprecation.
The block-based product editor is gone. The @woocommerce/product-editor package is out of core along with its routes, its feature flag and its extension points. Notice period was about five weeks between the announcement in late June and removal at the end of July.
Product data is untouched. This is the editing interface, not the storage. But if your extension hooked into it, those hooks no longer exist. Grep for three strings before anything else: @woocommerce/product-editor, product-block-editor-v1, and __experimental.
That third one is a good habit generally. Anything named experimental in a dependency can be removed without a deprecation cycle, and a periodic grep tells you what you are exposed to.
Three quieter ones that will bite
ReserveStock now defaults to a sixty-minute reservation window. If your checkout or inventory sync code assumed something different, it is now wrong in a way that manifests as overselling or as stock held far too long. Both are the kind of bug that surfaces during your client's busiest week rather than in testing.
Shop page queries changed type. get_queried_object() on the shop page returns a WP_Post where it used to return a WP_Post_Type. Any instance check or property read that assumed the old type breaks, and it breaks quietly.
The shipping-class taxonomy is becoming private, so audit anything querying it as a public taxonomy.
The performance work, and how much to believe
Twenty-eight pull requests in the release carry performance or scalability tags, concentrated where big stores actually hurt: the HPOS orders screen, catalog pages, the products list, point-of-sale catalog generation. Store API collection-data requests are de-duplicated to cut redundant round-trips, and product administration keeps persistent status counters instead of recounting on every page load.
Product object caching is on by default, though only for new stores. The quoted figures are around nine to twelve percent faster variable-product page loads and six to twelve percent on bundle checkout.
Treat those as a direction rather than a promise. They are vendor numbers with no independent verification, and caching gains are extremely sensitive to what is underneath. Benchmark your own store before and after or you are quoting somebody else's marketing back to your client.
A sane upgrade order
Install the WooCommerce Beta Tester and run the beta on staging. Grep for the three product-editor strings and fix what turns up. Audit the ReserveStock assumptions and shop-page type checks specifically, because those two are the changes most likely to pass testing and fail in production. Benchmark the orders screen if the store is large enough to care. Update the lint config while you are in there, since @woocommerce/eslint-plugin now uses standard import/order in place of the old dependency-group rule.
