Big Data Commences with Little Data
Any business that has been operating for a period of time is awash in data. Consequently, it’s probably safe to say “Big Data” is common to practically all businesses, regardless of size or age. But having more data doesn’t represent the answer to all questions since data in its raw form doesn’t reveal all that much nor is it all that revealing. This information only becomes interesting when you can transform data into intelligence, turning common and seemingly uncorrelated information into predictive insights. The possibilities are endless, making the allure so appealing.
There is, however, a fundamental flaw in the sequencing of this activity— the foundational work is often dismissed as boring and unimportant even though it represents the most significant work that must be completed. At New York Life, we are assessing the opportunities ahead of us and remain keenly aware of the appropriate sequencing and focus. Our approach is to commence with a “little data” strategy as a means to execute on our Big Data initiative. In other words, we are doing the foundational work—establishing organization, business objectives, governance, and data cleansing—before we are able to achieve full value from data. This foundational work is likely the first priority for most CIOs. Yet many organizations jump to more complex activities such as combining disparate data sources into a unified data store. This type of work is highly challenging and cannot be accomplished unless the foundational work is considered.
We have found the following to work well for us. Dedicate a team of professionals from both the business and technology sides to address data management.
While many firms are racing ahead to find and hire their Chief Data Scientist, they may discover that the need for such talent is not in the present but rather in the future. A well-resourced, highly collaborative team can attend to the basics, such as locating, cleaning and organizing data to enable a foundation from which to derive value. Consider leveraging resources from within your organization that understand the business challenges and data and co-mingle them with technologists who know the architecture and data structures. This bridge between the technologists and the business is critical to the success of any Big Data strategy.
Let the business drive the technology, not the other way around. The formation of any data strategy begins with identifying business objectives: what we are attempting to solve and why, what are the business issues that are the most complex, what are the most important problems and what offers the most P/L leverage? Again, this is best accomplished through a partnership of business and technology leadership. Insight into trends, anticipating issues and predicting from pattern recognition can only occur if everyone understands the business objectives at hand.
Create an organizationally appropriate data governance framework that fits with the culture of your organization. It is important to recognize how receptive your organization is to governance and create a framework that will be accepted and sustainable. Consider data stewardship. Having a single source of accountability or ownership for the data will contribute to a more mature and reliable capability. This includes processes and accountability for maintaining the data.
“While many firms are racing ahead to find and hire their Chief Data Scientist, they may discover that the need for such talent is not in the present but rather the future”
At New York Life, we are assessing how to ensure continuous improvement with regard to data governance and are evolving our thinking to achieve more informed business results.
Look for tools and accelerators. Although most actuaries and financial analysts have the innate ability to interpret data, translation becomes paramount for most others in the business. Therefore, tools need to be leveraged to help business leaders more easily understand and consume this information, and should include everything from dashboards to visualization. Proof of Concepts (POCs) should be encouraged that support and/ or accelerate this capability. With regard to the foundation, consider partnerships with firms that can automate the data “search and rescue” effort to accelerate the foundation building effort.
At New York Life, we have a technology roadmap that defines a series of initiatives aimed at progressing the organization’s capabilities in the data management domain. These initiatives include, but are not limited to: the creation of a data management function and governance processes ; guiding principles for investment; and technology solutions for reference architecture, master data management, data warehouse consolidation, tools simplification and advanced analytics.
Ultimately, data cannot become a strategic asset unless the foundational work is addressed. Harnessing disparate sources, establishing strong controls and standardization through tight governance, connecting technologists to the business leaders and partnering with industry specialists are essential elements for success. This represents a moment in time when the phrase “time to slow down in order to speed up” becomes the mantra required for companies to succeed in their approach to data insights and predictive analysis.