Panorama Business Intelligence

Using Advanced Insight to Drive Business Results

The business world is full of data and information.  Companies keep collecting it so they can better cater to customers, compete in markets and appease regulators.  Many companies are trying to adapt their business models and practices to integrate management by exception as a core principle.

The foundation of this management model focuses on the concept that a company has two types of products or processes – the ones that behave as expected and the ones that behave exceptionally.  Managing by exception is based on the premise that business information workers should be focusing their attention on the exceptional because that is where they can have the most impact on the overall business.  Business Intelligence supports and facilitates this management philosophy.  

Business intelligence tools are at the heart of enabling this philosophy. Nevertheless most BI solutions have become powerful in taking a large amount of data and presenting it nicely to business users. Yet, this has not helped in any way filter the data to truly focus on the areas that require attention.

In blogs that speak about BI many claim that BI has not proven to help the business become better mainly because there is simply too much data available.
Exceptions, as a way to filter the data, is available in business intelligence tools and enable companies to look at a large quantity of data in their operational data warehouses.  They can then filter that data in an intelligent way to highlight the exceptionally behaving and remove the data that is acting as expected. These capabilities enable business workers to focus on areas of the business that require attention or corrective action. 

While most BI solutions provide exception handling functionality, technology has not always supported managing by exception to realize its greatest potential. Companies need the flexibility and robust functionality to apply exceptions to whatever area of the business strategy dictates. 

However, most BI solutions available today only offer a modest level of exception functionality.  In a company already suffering from information overload, basic exceptions only compound the problem because results arrive with too much data to actually derive any meaningful insight. While management by exception is the ideal goal, the tools offer only basic exception handling capabilities.  This limits a company’s ability to manage by exception, making it challenging to drive the type of corrective action that improves business results.  

From a functional perspective, it is also not in the hands of the business users to create their own set of rules and exceptions and by that limiting the number of people actually taking advantage of that capability.

Limitations of simple exceptions
Simple exceptions limit a user’s data filtering capabilities.  For example, a user could filter data that is bigger or smaller than what’s expected in a specific data view. However, business users may want to build exceptions on data outside of what appears on a BI report. But simple exceptions do not enable users to create an exception on any other dimension other than what is available on a report. For example, if a user is looking at product sales by time, he may want to highlight the products that have low margins, even though margins do not appear on the report.

Simple exceptions also limit a user’s ability to chain exceptions. In most cases, users can only show one exception at a time.  As a result, a user can’t highlight and filter products that have a lower margin than product x, in addition to having the most defects. Also, when using simple exceptions, it is difficult for a user to set different exceptions for the same dimension.  For example, a user may want to set an exception where margins less than 45% in the USA and 35% in Europe are cause for concern.

Users need to see beyond data presented in a report.  Simple exceptions only focus on report data, which can sometimes be an aggregation of data. So, variances in product performance will not be readily apparent because an over performer could compensate for an under performer.  A typical exception management solution would not alert the user that further investigation is required.

BI Exceptions Explained
In the current BI landscape, exceptions generally come in two forms.  Users can either use basic exception functionality, which enables the creation of exceptions during data navigation.  The second form offers more sophisticated capabilities that reflect the natural flow of thought when reviewing data.  It also offers the ability to create exceptions on non- report data that using functionality including bubble up exceptions, which enable users to see unapparent data behaviors; chained exceptions, which enable a user to combine two or more exceptions for a rule; and custom exceptions, which enable analysts to write their own sophisticated exceptions using MDX.

For companies that are trying to adopt management by exception to handle information overload, basic exceptions can impede insight by limiting the types of exceptions as well as producing too much information.  For example, a company examining revenue or product performance may see satisfactory results when using a basic exception. 
However, what it may miss is the fact that one product is over performing while another is under performing.  The basic exception will show the balance of the two performance results without showing the underlying cause of that balance. 

Conversely, bubble up exceptions alert the user that exceptions exists a layer under the aggregated data and guides the user to identify the product that is under performing. Knowing about the discrepancies gives the company the opportunity to refocus its efforts on correcting the performance variations.

More advanced exceptions also facilitate the creation of more sophisticated comparison formulas.  As a result, companies can examine the exceptional at a higher level, moving beyond comparing larger and smaller percentages to analyzing factors such as dividing product revenue by sales representatives.

Radio Frequency Identification technology (RFID) is an excellent example of the power of advanced exceptions.  With RFID, every product in the supply chain becomes a piece of data available for analysis. Since there is so much data coming from RFID, users typically look at reports at an aggregated level and do not track each and every item on a single report.

Bubble up exceptions are extremely important in such cases because they alert users to products that may behave as an exception.  Users can learn this information without having to create a report that shows all of the data.  As a result, BI with bubble up exceptions supports supply chain management in helping discover trends about product development, fulfillment and execution. 

Ensuring correct execution and timely delivery – exceptions at work
A large US beer distributor wanted to measure the performance of its entire distribution business.  The company focuses on transporting product from ports to the warehouses and then to local and regional beer distributors.  Inside Customer Specialists interface with clients to fulfill orders and initiate the delivery process. 

The company developed approximately 60 Key Performance Indicators (KPIs) to measure the business processes impacting national distribution.  It wanted to track execution and delivery specifically.  With the KPIs in place, the company was able to leverage advanced exception functionality to pinpoint, for example, whether a customer specialist’s accuracy record was in line with established benchmarks. 

Using bubble exceptions, the company was able to identify, with minimal effort, where performance degradations were impacting overall business results.  As a result, the distributor transformed insight into action and focused on correcting the performance issues that were affecting its ability to achieve business goals.

In Conclusion
There is no dispute among businesses as well as BI solution providers that managing by exception is a critical component of any business management philosophy and, as a result, any BI system.  However, users experience limitations with many solutions because of the scope of the exception capabilities available.  Basic, user-driven exceptions are a good foundation.  But to become a company that derives business value from managing by exception, it is essential to have BI solutions with advanced exception capabilities. 

It is only with these types of capabilities that a company can transform itself into one that is truly managed by exceptions.  As a company managed by exceptions, it will know right away about the factors impacting performance.  Proactive insight will enable decision makers to take the required corrective action to improve business results.