I recently gave two presentations on trends in BI at IT conferences in Sweden and South Africa and one of the trends we are seeing very clearly is the trend towards enabling discovery:
Discovery (dis-ˈkə-v(ə-)rē): Find (something or someone) unexpectedly or in the course of a search. (ref. Google)
I recently gave two presentations on trends in BI at IT conferences in Sweden and South Africa and one of the trends we are seeing very clearly is the trend towards enabling discovery:
Discovery (dis-ˈkə-v(ə-)rē): Find (something or someone) unexpectedly or in the course of a search. (ref. Google)
To understand the reasons for this we need to understand why discovery is such an important aspect of the decision-making process that BI is supposed to support. As the definition states, discovery is about finding something unexpectedly during a search. While many discoveries are irrelevant and ultimately useless, others are so groundbreaking in their significance as to make all other useless discoveries worth the waste of time.
(Source: www.creativeclipsonline.com)
I wonder how many useless discoveries Sir Alexander Fleming made before he went on vacation in 1928, inadvertently leaving a petri-dish filled with bacteria to grow a mold that would ultimately be the genesis of the anitbiotic that has had such a profound impact on humankind over the last 80 years. The discovery of penicillin was neither expected nor even the reason why he was conducting his research. But it was discovered nonetheless.
The scientific world is full of examples of game-changing discoveries like Sir Alexander’s. Some have a smaller impact than his, yet are still very notable. From the discoveries that led to Post-It Notes, Velcro, cellophane, brandy (cognac), the microwave, Teflon, even Viagra: the world is full of examples of individuals or groups discovering something unexpectedly in the course of a search (often for something else!).
Discovery isn’t – and shouldn’t – be limited to the scientific community. Discoveries happen every day in the business and consumer worlds. In business, discovering that your company is selling more of a product in one region than another, or that on a given day of the week, at a particular time, the quality of product on the production line drops, or that when Product A is placed alongside Product B on your store shelves both products sell more – all of these have an extremely valuable place in our professional lives.
Up until recently, our business software has failed in its attempt to facilitate the discovery process. This is not an indictment of the people charged with implementing and running the projects; rather it is a reflection of the fundamental architectures that the traditional BI vendors were forced to choose when decision support became prevalent. Constrained by disk-based and query-based architectures because of the high cost of memory at the time, the natural free-flowing exploration of data that might ultimately lead to discoveries was not permitted to occur. In order to try to alleviate this restriction, the notion of OLAP and cube-based technologies came to the fore, but without the desired results. The fundamental restriction was the underlying reliance on disk-based and query-based technologies that were a) too slow and b) technically difficult to master. Ultimately, what BI became was the provider of standardized reports with well-defined KPIs and other metrics and dashboards which offered the users a limited data exploration experience.
It was here that “end users” truly became END users; they were the end of a process of data collection, modeling, and cleansing, and report generation. At QlikTech we talk about the “end of the end user.” With Business Discovery, the person actually using the software to make decisions through interesting discoveries is not thought of as the “end” of anything. Rather, they are the beginning of something. The beginning of the discovery process. The beginning of a process that allows them to ask and answer their own questions without restriction, without needing to follow pre-defined and restricted exploration paths.
The trend towards discovery in BI has been made possible with QlikView, which adopts a radically different approach to data access and analysis. The dramatic reduction of the cost of memory coupled with innovations such as associative data modeling and data exploration has given business users the ability to explore data to encourage those key discoveries. Gone are the restrictions around old, pre-defined data exploration paths, and are now replaced by an unrestricted approach to data discovery where business users can ask and answer their own questions, regardless of how the data has been structured further up the chain.
Imagine if Sir Alexander had been restricted by not being allowed to see the famous mold forming on his petri dish. In a traditional BI world, this path of insight would have been shut off to him, simply because it wasn’t something that had been pre-determined.