Data Granularity

The level of detail at which product or customer information is captured and stored. A catalogue with low granularity might describe a shoe with five fields; a catalogue with high granularity might use forty fields, including sole type, arch support, and intended terrain. Granularity determines the precision of every recommendation, filter, and upsell that the system can generate. It is the foundation beneath the recommendation engine, the search function, and the personalization logic. Investment in data granularity pays off most visibly in the long tail of the catalogue, where niche products can be matched to niche needs with accuracy.

Example: Netflix's 70,000 micro-genre tags allow it to recommend content along subtle dimensions like "quirky romantic comedies with strong female leads" rather than only broad categories.