1,000 parts to 64,000, still navigable.
A thousand products on WordPress to sixty-four thousand on Shopify, structured so a customer can still find one.
- 1k SKUs on WordPress
- Search the only reliable way to find anything,
- CATALOGUE ARCHITECTURE
- FILTERING
- PRODUCT DATA
- SEO
- SHOPIFY
- LARGE CATALOGUE
Context
Billy’s Speed Shop sells performance parts. When the work started the range was around a thousand products on WordPress. Eighteen months later it was sixty-four thousand on Shopify.
Growing a catalogue sixty-four-fold is a structure problem rather than a migration problem. A thousand products can be organised badly and still sell. Sixty-four thousand cannot. Past a certain size the only thing between a customer and the part they need is the way the catalogue is arranged, and most stores at that size are held together by search and hope.
The constraint
At sixty-four thousand products, inconsistency costs money.
A filter is only as honest as the data underneath it. If the same attribute is spelled three ways across the range, the filter quietly hides products that should match and the customer concludes you do not stock the part. If collections are built from the way suppliers send data rather than the way buyers think, people land on a page of nearly-right products and leave.
All of that had to be fixed without giving the team more admin, because a structure that depends on perfect manual discipline will not survive the next thousand products.
The approach
The work was deliberately unglamorous. Catalogue hierarchy, attribute discipline, filtering strategy and the SEO that follows from getting those three right.
Collections built around what a buyer is trying to do rather than how the supplier sends the file. Attributes standardised so a filter means one thing. Browse routes that reach the right part without the customer knowing the exact term for it.
Structure that holds as the range grows, rather than a rebuild that needs doing again at eighty thousand.
What we built
Catalogue architecture
A hierarchy built from buyer intent rather than supplier categories.
Collection structure rationalised to remove duplication and dead ends.
Rules for what belongs in a collection, so the structure stays right as products are added instead of needing policing.
Product data discipline
Consistent attribute patterns, so filters and search mean what they claim to mean.
Naming and specification conventions applied across the range.
The data edge cases that produce mismatched filters found and removed.
Findability
A filtering strategy that still works at full range depth, which is a different problem from filtering a thousand products.
Filters present where they help and absent where they crowd the page.
Browse routes to the right product for someone who does not know the exact term.
SEO
Information architecture and collection clarity as the foundation, rather than metadata laid over a bad structure.
Structural issues caused by thin and overlapping categorisation removed.
Operational fit
Every change weighed against the admin it creates. Structure that needs constant manual upkeep is not structure.
Results
The catalogue went from a thousand products to sixty-four thousand over eighteen months and stayed navigable. Customers could narrow down rather than guess, and the structure underneath held as the range expanded.
The work that matters at this scale is the part nobody sees. Clean attributes and honest filters are what let a store of sixty-four thousand parts behave like a shop rather than a warehouse.
From the outset, their professionalism and expertise were evident. Paul took the time to understand our vision, guiding us through every step of the process and ensuring our ideas were brought to life in a way that exceeded our expectations. The team's attention to detail, creativity, and technical know-how made the entire experience seamless and enjoyable.
Next steps
The engagement ran eighteen months, through the rebuild and then as ongoing work on catalogue structure, merchandising and search behaviour.