What is roll-up in data analysis?

Roll-up consolidates. Where drill-down expands a summary into detail, roll-up compresses detail into a summary along a hierarchy, days into months into quarters, products into categories, stores into regions into a company total.

Roll-ups are everywhere in reporting:

• Time hierarchies day to week to month to year.
• Geography store to city to region to country.
• Product SKU to category to department.
• Org team to department to company.

Doing this correctly means aggregating at the right grain so nothing is double-counted, a common source of error.

iDBQuery handles roll-ups from plain language and gets the grain right. Ask 'total sales by quarter' or 'headcount by department', and it aggregates your detailed records up to the level you named, writing correct SQL that respects the data's grain and relationships so totals are not inflated. Because you are querying live sources, you can roll up along any hierarchy your data supports, not just pre-defined ones, and switch between roll-up and drill-down freely in the same conversation. Every rolled-up figure is cited back to the detail rows it summarises, so you can always expand a total to see what is inside it.