Originally published August 19, 2008
Does your company have excellent management information? Congratulations! Your company is unique! Even though everyone acknowledges the importance of management information nowadays, why can’t we do it? Why don’t business intelligence (BI) and data warehousing initiatives provide what we all desperately want? In my view, good management information is important, but if we do not acknowledge and address conflicting interests, any BI project is likely to fail.
There are three parties involved in business intelligence, each of them having its own interpretation of good management information; however, their interests are quite different and even directly oppose one another. These 3 parties are:
When speaking of good management information, tactical/operational management inevitably associates good with fast. They want their management information as well as their management information applications as fast as possible. They want to act on their management information in daily operations because time is money. They typically act on the analyses of clever assistants and their home-built applications. Their databases have little data, little history and no documentation. MS Excel or MS Access can do the job. These assistants are highly appreciated within their organization, and they usually enjoy their status as a vital and often only source of management information. For them, it is a great feeling to know that the whole department depends on their magic tricks.
When the IT department speaks of good management information (in fact, any information), what they mean is a good information system – well managed, reliable and maintainable at low costs. They want secure systems that are well documented, with no double registrations or unmanageable applications. They want MS Excel and MS Access to be forbidden, and they hate it when departments ask for their assistance when one of their local applications crashes. IT will keep saying “I told you so.” In other words, they want to manage all data and all information, preferably with no changes whatsoever. Every change is a possible threat to the cherished equilibrium.
Corporate/strategic management mainly wants unique definitions. Reports that show costs should all have the same definition of costs. They want easy access and neat overviews over time periods (for instance, years). They hate the ongoing dispute about the actual meaning of numbers: What do they include? What is excluded? How can they compare the incomparable? What is the value of reports that show costs different from what the general ledger shows? They would rather have a single report than two reports showing different results.
Three parties, three grantable wishes – all worth doing something about. The interests are, however, quite different, if not opposing. What we have seen during the past years are projects initiated individually by each of these parties – resulting in failures. Different interests, different goals and different failures – but with one shared feature: they are expensive failures.
Projects initiated by corporate/strategic management often result in high consultancy costs and no managed process to actually produce the desired information. Strategic management hires a well paid consultancy firm that convinces the top management to do something about their key performance indicators (KPIs). The inevitable balanced scorecard and strategy maps are presented. Often held at exclusive resorts, these consulting sessions produce lots of paper reports outlining the corporate focus for the years to come. There are no instruments to actually control the process of measurement and evaluation against the targets, no tools to perform adjustments or corrections and measure again. Thus, corporate/strategic management will never really know whether the strategic adjustments have any effect.
Tactical and operational management are typically set in their ways. They have their accessible whiz kid, their fast Excel and Access applications. When needed, they will hire some more capacity to get their answers in time, thus generating hidden IT costs. The whiz kid spends 80% of his or her time on report building. The time spent will never appear as IT costs in the general ledger. The reports produced locally will have many equivalents in other departments that are more or less alike, thus creating a considerable redundancy in report-producing effort. All departments have their overview of personnel costs and productivity, overhead costs, etc. All of these overviews have their own implicit definition of those costs. When shown to the strategic management, these reports are like chalk and cheese. You simply cannot compare the incomparable
The IT department has some visionaries that read interesting articles about data warehousing. They present the management with the ultimate solution: a huge data warehouse, containing all the corporate data. All management information needs, now and in the future, will be covered since all the corporate data and the complete history will be ready and available in the data warehouse. All the data as well as the information will be managed properly with no redundant reports, no double definitions, and no Excel and Access dependencies. When the data warehouse is finally ready for use, it will be so heavily guarded that any possible user will lose his eagerness. What will happen is that operational systems will change and the data warehouse will not be updated. It will no longer have priority. Finally, the information will become unreliable. Unreliability will be the final blow to the ultimate expensive project. The IT failure is irreversible and no doubt the most expensive.
Good management information is achieved only by a combined initiative of all parties involved. All parties should be aware of the conflict of interests. These interests should be properly managed in a game of give and take, making full use of each other’s advantages: the whiz kid’s knowledge of data and business, the technical know-how of the IT department and the corporate views of the strategic management. Operational targets should be linked to the strategic goals. This will work. Projects initiated independently will not!
Note: This article is based on an article by the author that was previously published in Database Magazine, a magazine for professionals that is published monthly by Array Publications in Holland.
Recent articles by Karien Verhagen