In the beginning of this millennium the paradigm for a new Business Intelligence (BI) project was very different from today’s.
As BI consultant I was usually asked to build a new Data Mart, then design a large set of good looking reports with a few dynamic filters and guarantee they are refreshed every night. The main priorities for BI managers were: enhance visualization features, like filtering, slicing or drilling-down, not to mention pixel perfect design for printing purposes; exploring new delivery channels like mobile devices; improving OLAP cubes performance and capabilities as data volume growth is a concern; and assure that information is reliable and consistent with source systems – a common pitfall that still persists on BI projects that involve data integration.
Business Intelligence programs were still on early stages of maturity. I remember conducting a meeting to understand the reason why users were not using BI applications although their company made a huge investment on implementing them. Right after introducing myself and explaining the objective of that meeting I heard “someone already asked me about that last year!”. I had to overcome the surprise and ask gently to answer again, this time my questions. So, how often did he use BI applications in his job? “Just once, after the training session”. And why? “My boss told me to try but when I checked the reports it generated, the information was just not right”. It was easy to notice that no stakeholder was committed since the beginning for people motivation, and the lack of confidence in data was a major setback to the administration intentions.
But now, new challenges are forcing organizations executives not only to upgrade their BI platforms but also to rethink their vision. Leaders already started to see that People, besides having business and technological know-how, also needed to incorporate an analytical culture that could keep companies competitive in the market.
Analytics have changed the BI market landscape and decision makers have realized that velocity of data has increased. Operational systems save transactional data over which analytical models are applied, then instantly send alerts or provide information for managers to adjust actions. Social networks capture public reaction to products and campaigns, and in a few seconds information is ready to tell organizations if the impact is positive.
A mobile company may understand how the young public is reacting to a promotion offering summer festival tickets. An oil extraction company may discover a dangerous pattern on temperature level and take preventive measures. An airline company may detect a sudden peak of demand and adjust ticket price. A shopping center may offer immediate discounts when detecting a registered user mobile in its perimeter. A stock exchange operator wants to be on top of global news in order to invest or protect its capital. Doctors rely on clinical exams and data monitoring techniques to prevent infections on fragile babies. Authorities want to identify possible cyberattacks to keep functioning governmental critical systems.
The rate at which the decision agent must make choices is much higher than in the past. There are many situations where taking no action, or a late action, results in a competitive disadvantage or even a bigger loss.
This somehow reflects modern life, where everything happens and is announced in the next minute. Fast life, fast BI.
PEDRO MARQUES
BI & Analytics Consultant at Polarising