Decision Support System
For their tactical or their operational needs, companies need decision support system that provide information on all the metrics that govern their performance and the parameters that form the co-ordinates of their decisions. In the past, companies could barely integrate information from a few of their operational data stores and create standard performance reports to make decisions quickly; they looked for reporting tools where they could drill down and view from different angles. Increasingly, enterprises will need to inter-connect all their data sources into a decision support systems that take actions that are consistent with their enterprise strategy.
One instance of the use of performance measures is La Suisse Insurance which set targets for its sales force based on a multi-dimensional view of data. The company was losing up to $50,000 per salesperson each year by paying monthly allowances to salespeople who were underperforming. An OLAP tool helped in gaining multiple views of sales performance — by salesperson, branch, and region— which uncovered opportunities for raising productivity. Data warehouses can extend the capability and include the analysis of impact of prices, advertising, etc. on the results achieved.
Alternatively, companies can focus on decision-support tools which involve ad hoc queries and make considerably greater demands on the analytical and data management capabilities of business intelligence software. This kind of query requires much larger data sets for cross-referencing across several dimensions of data as well as over time. Companies need to be able to integrate data from several transaction data bases, usually in a data warehouse environment, and need the IT horsepower to conduct such complex queries. The typical applications of ad hoc queries are customer segmentation and response modeling.
Strategic decisions, unlike operational decisions, take a longer time as companies need to be able to parse current and historical data before they can come to decisions. They involve the use of statistical and data mining tools to consider alternative scenarios, predict future financial performance, conduct customer segmentation for product positioning and make decisions about the choice of their channels. This is typically done with data stored in data warehouses and updated periodically, typically overnight or on the weekends, in order not to interrupt analysis during the day.