Granularity (Records)
The degree to which objects are aggregated to records.
- •Ability to perform detailed, low-level analysis of transactions, events, or individual entities.
- •More accurate and nuanced understanding of patterns, trends, and anomalies.
- •Foundation for building more precise predictive models or detailed operational reports.
- •Flexibility to aggregate data to various summary levels as needed.
- •Inability to analyze data at the required level of detail, obscuring important insights.
- •Loss of information when data is pre-aggregated too early in the lifecycle.
- •Difficulty in identifying root causes of issues or understanding specific event details.
- •Limited ability to drill down into summary data for deeper investigation.
Story
Logistics: Gate transactions are recorded individually for each container move (in and out), allowing detailed analysis by hour, shift, day, or truck company.
Retail: Every click and page view on the e-commerce site is logged as an individual event record, enabling detailed customer journey analysis.
Manufacturing: Each individual product unit coming off the assembly line has its own quality inspection record detailing all measurements and checks.
Logistics: Sales data for container movements is only available aggregated at the monthly level, preventing analysis of daily operational trends or peak hour demands.
Retail: Website traffic data is only stored as daily summaries, losing information about user navigation paths or hourly visit patterns.
Manufacturing: Quality control checks are recorded only as a pass/fail summary per batch, not detailing individual item defects within the batch.