BigSmall. The filter rescue.
Their Shopify filters were dead on arrival: every tag empty, every SKU null. The fix started with finding out why.
The engagement. BigSmall runs corporate gifting on Shopify. Their catalogue needed reliable category and budget filtering, and every off-the-shelf approach assumed product data that turned out not to exist. Our job: find out why, then build filtering that works on the data that does.
Filters that assumed data nobody had entered.
Filtering a Shopify catalogue by category and budget sounds like a solved problem, and it is, if your products carry tags and SKUs. Every filter app and every standard integration leans on those fields. BigSmall's catalogue had neither: the tags were empty strings and the SKUs were null, across the board.
The tempting move is a heroic data-cleanup project: weeks of manual tagging before any feature ships. Tempting, expensive, and fragile, because untagged products would keep arriving.
Filtering on the structure that actually existed.
We root-caused the data first. The useful structure wasn't in tags at all: the team had been organising the catalogue into roughly 80 well-maintained collections all along, because that's how they actually think about their products. So we rebuilt the category and budget logic on collection membership, pulling live from the Shopify API.
No manual re-tagging project, no new data-entry habits for the team, no filter app fighting the catalogue. The system works with the grain of how they already work.
Working filters, zero cleanup project.
Category and budget filtering now runs against real, live catalogue data, and it keeps working as new products land in collections. The weeks of manual tagging never had to happen. No revenue claims here, that's not what this engagement was. It was a data problem, found and fixed at the root.
How this applies to you
If a feature keeps failing on your store or app, the bug is often in the data, not the code, and no amount of new tooling fixes a wrong assumption. Diagnosis first is the cheap path: it turns a weeks-long cleanup into a design decision.
That diagnosis is literally our Week-1 Build Audit: $500 flat, one week, credited against the build. Every blocker found or it is free.
Your project next