subscribe: Daily Newsletter


BI can learn lessons from the military (Part II)


There is a growing tendency for people to say business intelligence (BI) is overrated and has become a business commodity. However, if you effectively change your BI approach, BI will never become a commodity, writes Yolanda Smit, BI consultance at Harvey Jones Systems.

In Part I of this series, we started to investigate military intelligence as a discipline that serves as a good parallel for BI. Hopefully, we can learn some valuable lessons from military intelligence and start a BI revolution that will change the business value of BI forever.
A lesson learnt in Part I was that the “divine purpose” of BI is to be a central strategic tool of war. If BI is only used to automate mundane daily reporting tasks, then it is a definitely only a commodity. Furthermore, BI is important to all levels of the organisation, but the essence of intelligence required at each level will differ, as will the method of delivery.
This brings us to the next lesson: What can we learn from the three steps in the military intelligence process?
Often valuable information is gained from public sources and is freely available. This includes information such as population statistics, main industries, economic history, tonnage and basic weaponry. Even climate information, seasonal cycles, weather forecasts and lunar phases can be a goldmine of information that will give a country the edge in combat.
Next to gleaning intel from publicly available sources, the military intelligence also has to deploy more innovative strategies when collecting intel that might not be so freely available. This could be anything from planting spies to eavesdropping on radio messages or breaking encryption keys.
Although it might not be ethical for companies to resort to covert operations in order to gain business intelligence, there are still some key lessons to be learnt from the military’s approach to collecting intelligence. The art of collecting information is twofold:
* The first critical success factor is to ensure that you collect ALL relevant information that will paint a complete picture of the combat environment. A missing piece of information can lead to death, or even another 9/11. Companies may be a little parochial in the sources included in a business intelligence solution. Reporting on internal historical performance is not business intelligence. Only when this internal data is combined with available intelligence regarding the external environment and then aligned with the strategy of the company, does the information become intelligence.
* The second critical success factor lies in evaluating each available piece of data and information available, effectively evaluating it and filtering the noise until only the useful portions remain. Information and data are available in abundance, as can be proven by running a search on Google. Although supercomputers can store, process and analyse several terabytes of data might make it easy to analyse all available data, this is not a good BI strategy. Always remember the Pareto Principle and don’t waste 80% of your BI investment effort to gain only 20% value from the vast amounts of data available.
What is the value of data if it is not clear that it can be trusted? A vital function of the intelligence process is the analysing of data to ensure that it is accurate, relevant and valid. It is especially important to validate the human intel gathered by spies, which is notoriously prone to inaccuracy. Sources might just make up imaginative stories for pay, or might be motivated by a hidden personal agenda.
For this reason, no piece of intel can just be accepted as fact. It has to be scrutinized to validate the information and to assess its relevance to the current situation.
This is one step in the process of transforming data to intel that can be easily neglected in the BI process. The standard BI process tends to collect data and transform it into a format that is easy to use for business end-users. The BI process usually also test the accuracy of the data by balancing the data in the data warehouse back to the data in the original source system. However, the responsibility of validating the data and evaluating its relevancy is usually pushed back to the business end-users.
This can create a weak link in the BI offering as the majority of BI end-users might be middle- and supervisor-level managers who do not totally grasp the overall corporate strategy, and can therefore not assess the validity and relevancy of the data to the corporate strategy.
A BI solution that offers real business value should control all aspects of analysing data to ensure accuracy, validity and relevancy to the corporate strategy, before delivering intel to the end-user.
The final step in the military intelligence process is packaging the intel to make it easily available and understandable to end-users. Packaging combines two factors: delivery and presentation of information.
Information might be delivered online through an intranet with interactive capabilities for first-line intelligence personnel tasked with analysing all information in detail. They might require drill-down and cross-drill capabilities that enable them to investigate all possible threads of information to a possible root cause of a problem, threat or vulnerability.
However, the intel will not be delivered in the same format to the president of the country. On this, high-level information should be highly summarised and the critical components clearly highlighted.
Similarly, it is important for a BI process to deliver intel in various formats and through various media, whether through full-scale graphical OLAP interfaces with drill-to-detail capability, or a high-level executive dashboard. The method of delivering and presenting intel should be designed for each of the target audiences, so as to maximise the value added by the BI solution.
In summary, staying with the lessons learnt from the military intelligence parallel, we might need to ask several questions about the current BI process:
* Do we actually collect all the information required to paint an accurate, holistic picture of the internal and external environment of the company that provides the necessary intel for business users to make effective, value-adding decisions that enable the company to survive and progress?
* Do we invest enough effort to analysing and preparing all data collected to ensure that it is accurate, relevant and valid within the framework of the current environment and strategy? Or do we actually include irrelevant information that only serves to confuse and mislead business decision-makers?
* Do we effectively package the intel in a way that allows all the different consumers of BI to easily and effectively extract the value from the BI solution?
A BI solution that positively answers these questions above can never become a mere commodity.
Furthermore, this parallel study also implies that a BI solution can never be a once-off implementation that delivers the same information for all time. If BI is approached as a continuously improving process that evolves the through the cycle of collecting, analysing and packaging as the environment of the company evolves, BI can revolve back from being a commodity to becoming a core source of competitive edge.