Cube Analysis

Cube Analysis –or Cube-based BI tools provide simple slice-and-dice analytical capabilities to business managers. This includes:

• OLAP Functionality Against a Data Subset [cube]
• Pre-defined Analytical Views
• Slice and Dice Analysis

Prebuilt Cube Analysis allows business users to determine the root causes underlying data in reports, without the need to have skills for full ad hoc investigation of the databases.

Appropriate tools are a critical factor in any Performance Management System and using the right tool for the right purpose, is a key enabler. Cube Analysis is ideal for basic analysis that can be anticipated in advance, for example, analysis of of sales by region for certain time periods, or the analysis of sales by product and by salesperson. After reviewing initial reports, a sales manager may identify an issue, and use predefined cubes such as those mentioned above to investigate the root cause of the issue.

Root cause is typically found after referencing several pre-built cubes, with one or two cubes providing the context for a primary cube.

Once the cause of the problem is identified, most tools allow a link to the analysis cube can be sent to parties for further review and resolution.

By viewing a series of ‘report views’, using standard OLAP features [page-by, pivot, sort, filter, and drill up/down] a Cube of highly interrelated data can be viewed by various attributes defined in the cube [ stores, products, customers, suppliers] with any metrics in the cube [sales, profit, units, age], thereby creating various 2-dimensional views [slices].

Types of Cube Databases

MOLAP

Most OLAP vendors use custom-made proprietary cube databases [Multidimensional OLAP or MOLAP]. These cube databases have very small data capacities – less than 0.01% of real relational databases – however they are suited to the subsets required in departmental BI applications which are typically limited to between 10MB and 100MB of detailed and summary data.

Once a company needs to deploy hundreds of overlapping cube databases to cover
all the combinations of data subsets, summarization levels, and security privileges for different user groups
across multiple applications “cube farms” resulted, creating a drain on IT resources.

ROLAP

An alternate approach was developed by modeling the relational database as a “virtual multidimensional cube” with a technique known as Relational OLAP or ROLAP.

This enables OLAP against an entire relational database, without limiting what what data can be analyzed. The trade off is in slower response-times and overloading users with too many options or too much data, rather than a simple subset.

Intelligent Cubes™.

MicroStrategy resolved these tradeoffs with its Intelligent Cubes, providing functionality of small-scale MOLAP cubes with significant enhancements available only with a ROLAP underlying architecture. This overcame the speed constraints and allowed automatic ‘on the fly’ creation of cubes and manipulation of functionality such as filters.

NEXT: Ad Hoc Query & Analysis

Back To Top