Hey everyone, I’m working on a catalog-heavy app and trying to decide how far to denormalize product data across search, listing, and recommendation paths. Reads get much simpler when each surface owns its own copy, but every new copy seems to create another place where stale prices or inventory can leak through.
At what point does denormalizing shared product data become a long-term liability instead of a practical speed win?
Hari
