There is a great amount of agitation and investment around Business Intelligence by businesses across segments all over the world today. Wikipedia provides a definition of Business Intelligence as “computer-based techniques used in spotting, digging-out, and analyzing business data”. These systems are justified as measures for intelligent decision making. On the other hand, I am rather surprised that I could not find a corresponding definition for “Cognitive Intelligence”. The term cognition (Latin: cognoscere, “to know”, “to conceptualize” or “to recognize”) refers to a faculty for the processing of information, applying knowledge, and changing preferences. Cognition, or cognitive processes, can be natural or artificial, conscious or unconscious. Over the years as a Business Process expert, I have also seen a great amount of confusion differentiating between Data and Information.
The key operative word that springs commonly is “Processing” as data transforms to information and information to intelligence and intelligence to Innovation. At each stage, processing allows data to progress facilitating decision making that enables business performance as shown in the following illustration.
As is evident from the above illustration, while traditional business intelligence is restricted to processing of the data by the computer and is restricted to the first two stages of the process, cognitive intelligence is centric to all the stages in the above process and the processing is done by real people in their heads with the help of computers. Would it be unreasonable think that the benefits of all the investment in the so called business intelligence are worthless without paying equal or greater attention to cognitive intelligence?
I am not saying that I am the Einstein who has discovered this. But the fact that most organizations have business intelligence implementations which will never give them the business results they seek can be attributed to the following reasons:
- Enterprise Data is a derivative of an enterprise process. Where the processes of the enterprises are disjointed, the data can hardly be expected to be unified. Therefore, they would need to be extracted and treated (read staged) before Information can be generated.
- At the Information stage, if the roles of the enterprise processes are not defined, scenarios cannot be applied on the information generated and therefore no productive decision making can happen in a real-time environment.
- When process and roles are not aligned at the enterprise level, there is very little use of context servers continuously churning intelligence across scenarios; evidential or predictive.
Experts will tell you that when you buy a Business Intelligence solution for your enterprise, you are being sold exactly those three above points although they may be couched in infinitely other interesting ways.
It is essential to understand that Innovation that shapes action and the consequent results measured are what contributes to growth; the main purpose of calibrating business performance. These elements are experiential and human in their application and therefore cannot be automated. It is very similar to the how a human body functions. The sense organs only register a signal but it is the brain that processes the signal and produces intelligence that forms a collective response to the individual signals received from each of the sense organ.
Most enterprises that I have seen (and these range from some of the best Fortune rated organizations to start-ups) all have the same issues. They started with disjointed processes and continue to maintain islands of data and keep patching them with various technologies and gallantly continue to battle the patchwork to propagate intelligence post-facto for the demands of an online / real by the business –time engaging enterprise. While it may be difficult to dismantle years of data infrastructure overnight, even the emergence of Services Oriented Architecture (SOA) and supporting technologies have yet to see enterprises planning a departure from the past bad practices and modernizing their enterprise to respond in a unified manner to their customers. While Master Data Modeling (MDM), Unified User Profile (UUP) and Enterprise Service Registry (ESR) are just elements of the big picture and not the solution itself, the start should be with the unification of the enterprise processes, distinctly defined and treated for core and support; for the purposes of prioritizing and positioning the implementation and renewal of the enterprise. This will also pave the way to establishing and sustaining unified enterprise architecture across a global enterprise; tremendously reducing the cost of ownership of Information Technology & Systems as well as the implementation of new enterprise systems for strategic competitive advantage.
Unless the enterprise architecture aligns customers to the enterprise systems and infrastructure as delineated in the above architecture illustration, it will be impossible to derive the benefit of business intelligence specifically and information systems in general. This is true for any vertical and a few illustrative examples that follow will amplify the reality:
- In the retail sector where business intelligence is most employed and is desperately required in real-time, there is no use mining the transaction data alone. When there is a clear understanding of the decisions at the point to sale and loyalty (moments of truth), the process becomes renewing with respect to growth and managing customer / consumer aspirations & expectations concurrently.
- In the BFSI sector that also follows the employment of Business Intelligence in retail very closely, it is not the quantum of investments but the quality of investments that will allow the con figuration of new products and services to drive higher growth factors. It would mean selling coal to Newcastle if I were to suggest to financial experts that as the volume grows, the profitability reduces and this drop is usually supplemented by the premium from new products and services that a customer / consumer is willing to pay for a higher return on value.
- In the automotive sector, it is not just the buying factors but the influence of lifestyle that drive decision making for self, but the extended family and friends that needs to be understood to create lifetime value approach that guarantees growth. In essence, it is not the buying transaction but the life circumstances that surround a buyer that will provide the next sale opening.
The examples can go on but it is very easy to learn the fundamentals from the street corner vendors selling food. A lot of what they process and as a consequence store inventory and prepare food has got to do with the understanding of not how much they sell at a particular point in time or in a particular area but the type of food and the price propensity for quality that drives their business. If the shelf life of food is short, the shelf life of data and information is shorter. While it is not difficult to understand that the raw material for the final cooked food needs to be fresh, it is equally imperative in the case of data and information that needs to be processed for business. So any arguments against real-time data for business intelligence are a wasted effort.
Cognitive Intelligence which provides the experiential element to the raw business data is equally if not more important to make business intelligence relevant. It is even more important when businesses need intelligence in real-time to make decisions that can serve the purpose of transactional as well as the transformational. This can be only embedded in the enterprise processes and inculcated into the culture of the enterprise that implements expensive tools and gadgets to drive business performance. It is important to realize that the technology stacks available today allow enterprises to step change if they prefer as opposed to a big bang approach; transform the enterprise to a unified process mode of working. This is how growth and productivity can be sustained concurrently.
Cognition in simple terms is the ability to focus on the relevant (right type / kind) of data from a pile. As even neurosurgeons would tell you how a brain functions, this is what distinguishes a smart person. This is conditioned by the culture and environment that an individual grows in; however this is also a consequence of a number of hygiene factors such as diet and habits. In relevance to business intelligence, the cognitive intelligence has to be able to raise the right level of questions to investigate the available data and beyond to find strategic answers for business. This is governed by the kind of content exchanged by the enterprise. It would be very simplistic therefore to assume that business intelligence alone would provide the insights to business strategy. It would have to be a combination of Empirical and Analytical data subjected to the conditions of both business and cognitive intelligence. This is to a large extent the domain of Enterprise Information Management (EIM) that would need to be addressed in consonance with Enterprise Content Management (ECM). The following illustration that provides the framework for an Information Architecture needs to be considered when considering and / or implementing a business intelligence solution.
Merely seeing or hearing or tasting something has no effect for the human system. It is only when meaning is made to a certain relevance and / or context that the experience is fulfilling. Just like our senses, the enterprise systems may record the data but having the ability to interpret that “contextually” is cognitive intelligence and this is must be a critical element for a strategy to implement business intelligence and derive strategic benefits for the enterprise.