Arrival of the data warehouse application

Aubrey Van Aswegen, Data Futures GM at Software Futures, examines the emergence of industry-specific applications in the data warehouse market. It has been one of the great experiences of the data warehousing market over the last four years to witness how we have moved from trying to solve every business problem to focusing in on niches; to honing in on subject areas; to solving specific business problems. This trend has manifested itself in a particular way: today data warehousing applications are emerging to address the above requirements. This trend had a parallel in the online transaction processing (OLTP) world, where many applications sprang up over the years to support the transactional database. These applications focused on getting data in; the new range of data warehousing applications are focused on getting data out and solving specific business problems. Examples of this are to be found in retail, health and insurance, to mention just three vertical application areas. Application areas include customer profiling for customer retention programmes, risk analysis, shopping basket analysis and churn analysis, or the reduction of customer loss. Churn analysis is particularly important in the cellular technology arena, where 0,1% reduction of churn can translate to millions of rand in increased revenue. Data warehouse implementers are now seeking out-of-the-box warehouse solutions. Instead of spending months developing a data warehouse, the new wave of applications are executed in three steps: plug in, populate and go. There are many benefits for customers. The first is that the technology is hidden from view. This is the "black box" scenario, where the internals or underpinning technologies are of no interest to management. All they know is that they`re getting an easy-to-use data warehouse application that plugs into existing data sources, transforms the data into information and is presented to businesspeople in a way that is easily understandable. And it is delivered in days or weeks rather than months or years. Under the hood of the data warehousing application are all the traditional disciplines associated with data warehousing: cleansing of data, maintenance of referential integrity, provision of application-specific source data for the data warehouse, aggregation of data, and definitions of the most common queries for ease of use. But the business is not concerned with these issues; all it sees is a functional data warehouse that is delivering benefit. And, as that is what business has sought from data warehousing all along, it is a trend we are heartily embracing.