Custom Software Development

Business Intelligence

Introduction

Business Intelligence (BI) takes its name by analogy from military intelligence. This is a pity, because military intelligence tries to find out something useful about other people or organisations, whereas BI refers to what you can and should know about your own organisation, and is not to be confused with commercial espionage.

The abbreviation BI is but one acronym in a field awash with the things, including ECM, BPM, KRA, EII, KPI, ERP, CPM, EAI, ETL, DW, BAM, ODS. Enough! This may all be sustenance for the IT super-savvy, but let's come down to a definition mere mortals might find useful:

Business Intelligence is everything you need to know about your business's status and performance, when you need to know it and in the form in which you need to know it.

History

The purpose of BI is to underpin, inform and guide decision-making. There are commercial systems available to help companies achieve this simple-seeming end, but these have been way too complex and expensive for all but the largest companies and corporations. Business establishments have always generated masses of information and data, and the problem has been how to bring all these together into a quickly informative database.

Historically, and until quite recently, most BI applications, such as decision support systems (DSS’s), were entirely data oriented, e.g., sales conversion rates, inventory turnover rates. There is a serious problem when dealing exclusively with quantified data, a difficulty also experienced by disciplines such as education, psychology, demographics and economics. It is that numerical data, no matter how relevant and 'accurate', always mask the reality behind the numbers. If the numbers tell us that we made x% profit on y sales, we learn nothing about the resources and effort that were required to bring about those sales and that profit – what helped and what hindered.

This inconvenient truth led to the eventual focus on Key Result Areas (KRAs) and, later, Key Performance Indicators (KPIs). The KPI Scorecard, together with any number of Work Performance Management (WPM) systems, became a useful process by which any organisation could keep tabs on who is doing what, and with which and to what effect. Yet even these approaches have remained limited by a reluctance, on the part of decision-makers, to step outside their comfort zone of hard data. Most KPI scorecards still omit measures, no matter how tentative or crude, of actual relevant work performance such as know-how, motivation, morale, and of the company's HRM mechanisms designed to support such aspects of performance.

Trends

Several major trends in BI have been identified. These are:

• Single purpose vendors, e.g., providers of data quality applications, are being gobbled up in vast numbers by larger providers of the complete ETL package – Extract relevant data, Transform it into a useable form, Load it into the appropriate database or warehouse.
• These newly expanding BI providers are also taking over or forming partnerships with EII (Enterprise Information Integration) vendors, greatly enhancing real-time data integration capabilities.
• These bright new and innovative companies are then snapped up and their skills and software incorporated, mainly by IBM and Oracle, which then provide what is becoming known as master data management (MDM) applications.
• The historical BI focus on strategy and back room activities in general is shifting towards incorporating BI into operational areas. There is a new emphasis on reducing the time taken for many tasks: preparing data, analysing operational data, and making decisions and carrying them out as a result of the data provided.
• There has been a remarkable growth in the sophistication and the user friendliness of predictive analytics and software enhancing and guiding decision-making capabilities.
• BI data will always be important, and now people are coming to realise that BI content, which essentially means anything alpha rather than numeric, must also be built into any comprehensive system.

How to do BI

BI is essential to the success of all businesses, from sole traders on upwards. The CIO Business Technology Leadership organisation presents an excellent ten point summary of how to go about identifying exactly what an organisation needs in order to enhance its BI to the maximum. Here is a summary of their suggestions:

1. Sponsor. Choose a sensible, broadly informed highish-level and well regarded executive, not in IT, as BI sponsor, someone familiar with at least the concept of translating the company mission into key performance indicators (KPIs).
2. Create common definitions. Everybody must mean and understand the same thing when any term is used or discussed, e.g, gross profit margin, work performance. Initially, keep the number of KPIs small.
3. Assess the current BI situation. Do not skimp here. Know what you know, thereby helping to clarify what you don't know.
4. Create a data storage plan. Onsite or virtual? Traditional data warehouse or cutting edge metadata repository?
5. Understand what users need. Be very clear as to what your strategic, tactical and operational staff require by way of information.
6. Buy or build the data model? Probably buy if you have one ERP and one CRM. Out-of-the-box, industry-specific models are available. Keep in mind you want to be able to increase scale over time and to extend to include more wide-ranging data. This could imply build rather than buy.
7. Consider all potential BI components. You don't want to find yourself bemoaning inadequate data quality, analytics or presentation media.
8. Choose a systems integrator. Be prepared to pay $5 to $7 on specialist services for every $1 spent on software. There must be close collaboration between end users, analysts and developers.
9. Think 'actionable' and 'baby steps'. Choose an end user, analyst and developer to create a first proof, using a few KPIs to develop a few reports.
10. Choose low hanging fruit to start. Start with high value, simple components. For example, a sales analytics presentation will present high value targets in an area of activity where there are many existing models and best practices available for comparison purposes.

One final point. Your workforce knows more about your business and what you need to know than will any consultant or software. The platitude 'Our people are our greatest resource' is nowhere more true than as a source of the information you need. Choose a low ego consultant to assist, someone who wants to go beyond simply vacuuming staff knowledge and feeding it back to you. Seek a facilitator, someone skilled at bringing top management and staff together.