SEO controls
Filtered URLs are useful for shoppers, but they can create too many crawlable combinations if every price, colour, size and stock state becomes a separate indexable page. Sieve PRO is planned to add a controlled SEO layer for stores that want useful filtered landing pages without generating thin search results.
Planned controls
Section titled “Planned controls”| Control | Purpose |
|---|---|
| Canonical strategy | Point low-value filter combinations back to the base category or shop page. |
| Noindex rules | Mark selected filter patterns as non-indexable. |
| Indexable combinations | Allow only approved combinations, such as category + brand or category + material. |
| Landing page metadata | Set title, description, H1 and intro copy for curated filter states. |
| Sitemap inclusion | Include only approved landing pages, not every possible query string. |
| Robots hints | Keep filtered URLs shareable while limiting crawl waste. |
Example use cases
Section titled “Example use cases”- Index
/shop/shoes/?sf_brand=acmeas a brand landing page. - Keep
/shop/shoes/?sf_price=10-20&sf_stock=instockout of the index. - Give
sofas + leather + beigea custom title and description. - Prevent filtered pagination from competing with the canonical landing page.
URL behaviour
Section titled “URL behaviour”Sieve already keeps filter state in the URL. PRO SEO controls should work with that model instead of replacing it. The shopper can still share filtered results, while the store decides which of those states deserve indexation.
Implementation notes
Section titled “Implementation notes”The SEO layer should run after filter state parsing and before page metadata is rendered. It should be deterministic, cache-friendly and independent from the AJAX endpoint, so the initial server-rendered page has correct metadata.