Traditional BI Process

Historically, Business Intelligence (BI) has been about a report-centric architecture that’s IT-driven and highly controlled. IT departments take operational data, create data models and build reports for end users. Traditional business intelligence delivers predefined reports and queries and pre-configured dashboards.

Let’s look at what’s involved in a traditional business intelligence process. Initially, a data warehouse would be set up to consolidate data from many sources. Data warehouses are designed to run analysis on large volumes of historical data. To get the data into the data warehouse there is then an ETL – Extract, Transform and Load process that it needs to go through. The transform step can include multiple data manipulations such as splitting, translating, merging, sorting & pivoting. The final Load step involves loading the transformed data into the data warehouse.

So in traditional BI, data warehouse and ETL tools support “back end” processes. The next step is usually to use an Online Analytical Processing (OLAP) tool to support the presentation layer or “front end” processes such as querying, analysis and reporting. OLAP tools allow multi-dimensional data analysis & the capability to present this in a visual interface with charts, tables and graphs. Cubes are the core components of OLAP systems; they aggregate facts from the data at multiple levels.

With traditional BI, data sources are normally pre-determined by IT.  End users don’t get to interact with the data in the systems. The analysis and reporting is generally conducted by IT. The analysis and reporting  is hard-wired into the data warehouse and cube structure. Each time data changes are required, the data has to go back through the time consuming ETL process described above, which could involve analysis across networked databases and hard drives. This can be a very slow process.

Business Discovery Process

With Business Discovery Tools like QlikView, this process is very different and much faster. QlikView takes all of your data from virtually any source and consolidates it into a single data pool. Next, QlikView loads the entire data pool into memory. Users don’t have to pre-process data and organise it into cubes and there are no hard disks involved like in traditional BI. Accessing data in memory is much faster than accessing data from disk.

With QlikView users get to search and analyse all of their data in one place. As the data is in memory, there is no need to wait on data refreshing with each new data selection. Users can instantly select data they are interested in, make associations and see relationships between data residing in different applications. They can perform ad hoc queries on the data and ask ‘what if’ type questions all within the one application. Business discovery tools like QlikView are fast, intuitive and simple to use. They give users the data that they need at their fingertips.

Business Discovery is about self-service Business Intelligence. It enables business users to customize and change analysis to address their exact business problems. Business users get access to the data they need when they need it and they get it fast!

Article by Carolyn ClarkCarolyn Clark is Business Intelligence Account Manager at Eureka Solutions and specialises on QlikView Business Intelligence software. Carolyn has a degree in Management Science and has many years’ experience in the software industry focussing on business process management & business insight products. To get in touch with Carolyn, email carolyn.clark@eurekasolutions.co.uk or call 01355 581 960.