Analytic applications

Analytic Applications are a type of business application software, used to measure and improve the performance of business operations. More specifically, Analytic Applications are a type of Business Intelligence solution. As such they use collections of historical data about business operations to provide business users with information and tools that allow them to make improvements in business functions.

The maturity levels for business intelligence solutions are as follows:

It may extend further to predictive analytics, or predictive analysis may form part of the analytic application - depending on both the subject matter under analysis, and the nature of the analysis required.

Analytic Applications are typically described as a subset of performance management. They specifically relate to the analysis of a business process (such as sales pipeline analysis, accounts payable analytics, or risk adjusted profitability analysis) in support of decision making.

To qualify as an application (rather than simply as a data warehousing tool), these tools should promote some form of automation. The maturity level of this automation is as follows:

reading data from a nominated operational system (ERP, CRM, SCM, etc) into a data warehouse optimized for analysis (data led automation),
reports, dashboards and scorecards based on that data structure (reporting led automation),
what-if analysis and scenario-modeling (predictive or analytic led automation).

In most cases, these three levels are discreet functions, loosely banded together as a single product, and there is little automation of the process from end to end.

Implementing an Analytic Application is not without issues. You may wish to consider the impact of customizations to your source systems, and to the re-work that may be required when you upgrade those source systems. In many cases this customization will require 20% or more of your "basic" or "vanilla" application to be reworked. In some cases this will be significantly more. Accurate project scoping is essential.

Anticipate that you will encounter data quality issues. These are often masked by the logic built into your source applications, and in many cases, cleansing or resolving the quality issues will double the initial project timetable.

See also