OLAP (Online Analytical Processing) is a data-processing approach optimised for complex, multi-dimensional analytical queries — 'what is the margin by customer, by product, by month, compared with the same period last year'. The opposite is OLTP (Online Transaction Processing), optimised for fast individual operations ('record this sale').
OLAP systems organise data into multi-dimensional cubes (time, product, geography, customer — each a dimension) with pre-calculated aggregations. Queries that in OLTP would take minutes or hours respond in seconds. The price is additional storage and the need for a periodic refresh when the source data changes.
Modern OLAP engines (including Qlik Sense behind MyBusiness-ITV) implement this paradigm without requiring rigorous pre-definition of cubes — the user explores dimensions freely, and the engine optimises queries on the fly. Qlik's associative model is especially powerful for OLAP analysis in cases where the relationships between dimensions are not fixed, such as in retail with multi-channel customers.