# Online Analytical Processing (OLAP)

> Big queries that add up lots of rows. It is the work behind reports.

**Field:** Data Engineering  
**Category:** Foundations  
**Also known as:** OLAP, Analytical Database

## Think of it like

It is like adding up a whole year of till slips. You want to see which month sold the most.

## Definition

OLAP is the kind of work done to answer questions about lots of data. A query reads millions of rows but only a few columns. It adds up, counts and groups them, such as sales by month and region. These queries are fewer than OLTP ones, but each is much bigger. Systems built for OLAP usually store data column by column.

## Example

A manager asks for total sales by country for the last three years. The query scans 200 million rows and returns 40 numbers.

## Why it matters

Reports and dashboards run on it. A system built for OLAP can answer in seconds what an app database takes hours on.

## Common mistake

Using it as an app's main database. It is slow at many tiny writes.

## Three principles

- **Big scans.** Each query reads many rows.
- **Few columns.** Only the needed fields are read.
- **Column by column.** Data is stored by column for speed.

## Related terms

- [Online Transaction Processing (OLTP)](https://learn.benibix.com/term/data/oltp)
- [Hybrid Transactional/Analytical Processing (HTAP)](https://learn.benibix.com/term/data/htap)
- [Denormalization](https://learn.benibix.com/term/data/denormalization)
- [Database Query](https://learn.benibix.com/term/development/query)

## Told apart from

- [OLTP vs OLAP](https://learn.benibix.com/compare/oltp-vs-olap): People think OLTP and OLAP are two brands of database. They are two kinds of work. OLTP is many tiny reads and writes, such as placing an order. OLAP is a few huge queries that add up millions of rows, such as sales by month. A system built for one is slow at the other.
- [OLAP vs OLAP Cube](https://learn.benibix.com/compare/olap-vs-olap-cube): People think OLAP means a cube. OLAP is the kind of work: big reads that add up many rows to answer questions. A cube is one old way to do that work, by adding up every total in advance. Most OLAP today runs in column-based warehouses with no cube at all.

## Taught in

- [OLTP, OLAP and HTAP](https://learn.benibix.com/learn/data-foundations/oltp-olap-and-htap): Data Engineering Basics, 11/11

---

Source: https://learn.benibix.com/term/data/olap
Author: Behnia Zarei (a developer in Vienna, Austria, who builds and teaches data, AI and business intelligence at BENIBIX)
Publisher: BENIBIX Learn (https://learn.benibix.com/)
Last updated: 2026-11-20
Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Quote freely with attribution.
