Saturday, February 21, 2015

Digital Master Tuning XXXXIIII: How Can Big Data Make the Organization Digital Fit

A company that is dysfunctional in digital analytics is going to struggle to make the transition to the digital paradigm.
Big Data and Predictive Analytics are at the top agenda in the majority of organizations. However, most of businesses are still struggling to provide their users with timely answers to new and changed requirements, relevant, consistent information of sufficient quality at acceptable costs. How can Big Data make the organization digital fit from getting fundamental right, and ultimately grow into Digital Master?

Data baggage symptom:  Many organizations are indeed struggling to provide the right information to the right people at the right time, but the same organizations are also looking at new initiatives like Big Data and Predictive Analytics in the hope that they will somehow be easier and will make their users happy. Why are there so many businesses can’t make the basic right? That's because most organizations come with data baggage or some call it the "databesity", which is what costs time and money in terms of integration, performance, risk management, data governance, data lineage etc. The foundation of getting the Big Data lies in keeping the basic right. Gathering the relevant, consistent and actionable information lies in business processes and structured methodology. A lot of information will end-up being relational as opposed to transactional, and making sure that sanity checks are done regularly on information gathered by varying resources is crucial.


Data Quality is a big issue: In most companies where the data quality is poor, it has to do with definitions, if you don't know what you mean by "profit"or "cost price," and you let everybody define it their own way in their own spreadsheets, then you are never going to produce information that makes any sense -It doesn't matter if it comes from Big Data or Small Data. The data blending facilities in some of the new products don't help the information quality much either - they create their own version of virtual reality. And it would help if the CIO would take up the responsibility for the design, implementation and exploitation of a value-added information management process based on a solid framework, as exist for the logistics of "non virtual" assets. This is the only way to avoid silos. But with all the analytical capabilities that current technology has enabled, corporations would increasingly place their decision making on what these analytical tools come up without applying some good old managerial judgement.


Data governance as part of business governance. Data governance (master data definitions, business rules definitions and data quality rules) would be a great help, but as long that is done whilst extracting, blending and integrating data "between source and BI," it remains carrying water to the sea. Unless the business initiates real support and takes the responsibility for data governance, the situation will remain as it is. And on top of that,  there are way too many IT people doing BI, who don't understand the first thing about BI. Many organizations deal with compliance-related data, captured through a system put in place over 10 years ago, when the focus was on automating workflows through web interface. At the time, reporting requirements and expectations were minimal compared to what they are today, not to mention that the regulatory burden keeps growing. Starting small, by examining a subset of data elements will help answer relatively simple but pertinent questions around compliance and process performance. The idea is to develop a bottom-up pilot project, starting with developing metadata and ending with reliable reporting. And it will be done with the tools at hand until you understand the path forward well enough to make a business case for scaling up.


Quick wins: In order to make Big Data more “visible” for shareholders, the companies need to investigate areas where they can have "quick wins" with new approaches of big data. Targeting customers is such an area where it is relatively "easy" as it - in general - does not require an integrated 360 degrees approach over all data sources as BI for a proper execution of the planning & control process to support strategic and tactical decision making does. However, the balanced analytics approach by focusing on long term goals with some quick wins is optimal on the Big Data journey.


A company that is dysfunctional in digital analytics is going to struggle to make the transition to the digital paradigm. These same organizations typically don't have good governance or data quality processes in place that a more mature analytics practice would. Or it will fail to develop them during their Big Data journey. Some companies will learn and adapt. But starting the journey is not contingent on having established processes, if they are not solidified in time then the likelihood of reaching the massive potential of Big Data will diminish to the point of no return. What can be done? The more that the enterprise adopts Big Data via logical steps and talented analysts,  the more Big data practices will mature.


Digitalization is like a flywheel, and Digital Masters are the one riding above it. Surf more Information about Digital Master:



0 comments:

Post a Comment

Twitter Delicious Facebook Digg Stumbleupon Favorites More