Article

How I Manage Product Data Across 200,000+ Ecommerce SKUs

The practical patterns I rely on to keep a large, multi-source catalogue clean, current, and trustworthy.

The short version

Most of the pain in a large ecommerce catalogue isn't any single hard technical problem — it's the accumulation of a hundred small inconsistencies across suppliers, formats, and time. Here's the set of patterns I've found actually hold up at 200,000+ SKUs.

Treat every source as untrusted

It doesn't matter if a supplier feed has been reliable for two years — the day you stop validating it is the day a malformed export takes ten thousand SKUs to zero stock. Every ingestion job should assume the incoming file could be truncated, reordered, re-encoded, or just wrong, and validate before anything downstream trusts it.

Matching needs a confidence score, not a boolean

Product matching — deciding whether a new record is the same as an existing one — is rarely a clean yes/no. Exact key matches (a stable part number, a UPC) are easy. Everything else is a confidence problem: title similarity, brand plus attribute overlap, price plausibility. The systems that work well score that confidence and route anything below a threshold to a human, instead of forcing a binary decision the data doesn't support.

Matching confidence routing

Exact key match

Auto-approved

High-confidence fuzzy match

Auto-approved, logged

Low-confidence match

Human review queue

Diff before you publish

The single highest-leverage habit in catalogue automation is generating a diff report before anything goes live — new items, removed items, changed fields, all visible before publication. This catches more real problems than any individual validation rule, because it surfaces things you didn't think to write a rule for.

The hardest part isn't technical

The genuinely hard part of running a catalogue this size isn't the code — it's the organizational discipline to keep supplier relationships, data contracts, and escalation paths clear enough that the automation has something reliable to build on. I go deeper on the system itself in the catalogue automation case study.