By Dewald Opperman,
Business Intelligence (BI), in general terms, is the process of turning raw data into meaningful information that a business can use to its advantage, using business processes, methodologies and technologies. This may sound straightforward, but getting to that meaningful information can tell you many things about a business organisation. Businesses often say “We need improved BI” but, having apparently improved their BI, continue to complain that they can’t seem to get reliable information out of their systems or that they do not know what to do with the information. This raises the question of the scope of BI and the approach to really make it work. Is only about extracting, presenting and analysing data or does it extend beyond that?. If the hype about the term when it first became popular was to be believed, BI would help business users make better decisions by producing a “single version of the truth”. It would enable businesses to make the right decisions faster and help companies respond to opportunities while those opportunities were developing.
Some of the challenges in achieving an optimal state are illustrated by actual examples. In one particular case, an attempt to implement an improved BI system revealed that the company’s implemented ERP system did not have effective reporting, and basic metrics could not be accessed by business users. Upon delving deeper, it was discovered that some processes in the business were run entirely off MS Excel. While BI can tap into these “data sources”, it is expected that it can also deal with the fact that these very Excel sheets are changed on the fly every month to cater for some obscure, unique condition. A slightly unrealistic expectation, especially while the alternative, more formal system exists already.
In another case, a business had implemented a BI system: They had dashboards deployed over a portal and accessible to internal managers and external clients. The right metrics were all there and these were refreshed mostly on time to aid operational and tactical decision making. Their complaint? Managers in their business weren’t looking at the dashboards, because it was so often inaccurate that they preferred to rely on information they had gathered themselves. On investigating the root causes of the problem two issues were discovered:
- Some automated data feeds would sometimes contain incomp.lete data
- System-based operational processes were sometimes not followed, leaving gaping holes in the data available to monitor the end-to-end process. Again, the initial BI development was sold on the promise of “operational visibility” and empowering managers to make decisions based on real information.
For both of these cases, two key components were missing:
- Process discipline and ensuring business processes are followed
- Monitoring and acting on data deficiencies, in close to real time.
Both the businesses concerned were surprisingly close to getting value from BI. Underlying systems were in place, reporting systems and supporting data structures already existed. This is one of the most exciting things in the area of BI today: systems and tools are available in abundance and often key components have been implemented already. In addition, products have been developed to service a need for visual analytics, making it easy to integrate data from various sources, graphically show deficiencies and, ultimately, get to useful information.
When businesses ask for help with BI, they believe they know what they need to measure to improve their organisation, but believe they need help in understanding what data elements are needed reliably and consistently.
The reality that such businesses need to understand is that this is not Business Intelligence per se, but rather a form of sourcing monitoring and process adherence diagnostics. BI today is elevated beyond such somewhat narrow technical definitions, and starts encompassing business process work as well as organisational change work. As we can see in both case study examples, behavioural change in the organisation is often the key to unlocking valuable BI from existing business data.