OLAP — Online Analytical Processing
OLAP (Online Analytical Processing) describes analytical-database technologies and methods for complex multi-dimensional queries against ERP and warehouse data. OLAP emerged in the 1990s as the analytical counterpart to OLTP (Online Transaction Processing — the transactional ERP world). For DACH ERP-bearing organisations, OLAP underpins much of business intelligence and management reporting, though the terminology has shifted toward 'analytics' and 'BI' in modern usage.
Multi-dimensional analysis
OLAP organises data as cubes with dimensions (time, customer, product, region) and measures (sales amount, quantity, margin). Users navigate the cube through slice (filter by one dimension), dice (filter multiple dimensions), drill-down (move to finer granularity), roll-up (move to coarser aggregation) and pivot (rearrange dimensions). The cube model fits intuitively how managers think about business performance — 'show me sales by product and region for the past 12 months' is a natural cube query.
MOLAP versus ROLAP versus HOLAP
Three architectural patterns. MOLAP (Multi-dimensional OLAP): data stored in optimised multi-dimensional structures (cubes). Fastest query performance, requires pre-aggregation, limited flexibility. Tools: Microsoft Analysis Services Multidimensional, Oracle Essbase, IBM Cognos TM1. ROLAP (Relational OLAP): data stored in relational databases with star-schema modelling; OLAP queries dynamically generate SQL. More flexible, less performant, scales to larger data volumes. Most modern BI tools (Tableau, Power BI, Qlik) are essentially ROLAP. HOLAP (Hybrid OLAP): combines MOLAP-cached aggregates with ROLAP for detail-level queries. Modern in-memory analytics (SAP HANA, Microsoft tabular models, Snowflake) blur the lines between these patterns — the speed of modern columnar storage often eliminates the need for pre-aggregation that defined classical MOLAP.
Leading OLAP and BI platforms
Modern BI (largely ROLAP / HOLAP): Microsoft Power BI (with Analysis Services tabular models), Tableau (Salesforce-owned), Qlik Sense, MicroStrategy, IBM Cognos, SAP Analytics Cloud, Oracle Analytics Cloud, SiSense, Looker (Google). Classical MOLAP: Microsoft Analysis Services Multidimensional, Oracle Essbase, IBM Cognos TM1 (now Planning Analytics). In-database analytics: SAP HANA, Snowflake, Databricks, BigQuery enabling SQL-driven analytics without separate OLAP layer. Specialist financial OLAP: Anaplan, Workday Adaptive Planning, Jedox (DACH-built planning platform), Oracle EPM Cloud, OneStream. For DACH mid-market, Microsoft Power BI is now the dominant BI tool, often complemented by financial-planning specialists like Jedox or LucaNet for budgeting and consolidation.
Practical considerations
Three patterns shape ERP analytics in 2026. (1) Embedded analytics: SAP S/4HANA Embedded Analytics, Microsoft Dynamics 365 with embedded Power BI, NetSuite SuiteAnalytics provide in-context reporting without separate data warehouse. Suitable for operational reporting on current ERP data. (2) Cloud-warehouse analytics: ETL extracts ERP data to a cloud warehouse (Snowflake, BigQuery, Databricks) where Power BI or Tableau analyse it alongside data from other sources. Suitable for cross-system analytics and historical analysis. (3) Real-time analytics: HANA-style in-memory analytics running directly on the operational ERP database eliminate the ETL step for time-sensitive use cases. The right pattern depends on data-volume, latency-sensitivity and analytical complexity.