Shopify SEO, from the structure up
Shopify SEO written from the technical side: collection architecture, product schema, faceted navigation, and getting cited by AI search.
Most Shopify SEO gains are structural rather than editorial. How collections are built, which URLs your filters generate, whether product schema matches the page — organic traffic largely follows from those three decisions.
These articles cover that ground in order: structure first, then measurement, then the newer surface of answer engines.
Articles on this topic
Shopify collection architecture and organic traffic
Most search demand is category-level, not product-level. How to build collections for search rather than for the navigation menu.
Product schema on Shopify: what search engines actually read
Most Shopify themes emit product structured data, and most of it is subtly wrong. What the markup needs to contain, the errors that suppress rich results, and why answer engines raise the stakes.
Filters and faceted navigation: the quiet crawl budget leak
Shopify filters generate near-infinite URLs. Which ones deserve to be indexed, which should never be crawled, and how to tell the difference without guessing.
Getting a Shopify store cited by AI search
Answer engines read stores differently from search engines. What makes a page quotable, why structured data matters more than it did, and which habits actively exclude you.
Frequently asked questions.
Where should Shopify SEO work start?
Collection architecture. Most search demand is category queries rather than product names, so if collections were built for the menu rather than for search, that demand has nowhere to land.
What is the most common technical SEO mistake on Shopify?
Letting filtered URLs be crawled without limit. Four filter groups on one collection produce hundreds of addresses, and crawl budget goes to colour-and-size combinations instead of new products.
Is valid product schema enough for rich results?
No — it also has to be true. Markup saying £40 while the page says £35 is valid and wrong, and that kind of mismatch undermines trust across the whole catalogue.