Emerging Trends in Predictive Analytics
Date: Friday , January 10, 2014
Headquartered in New Jersey, Dun & Bradstreet (NYSE:DNB), a leading source of commercial information and insights on business, is serving the global customers for more than 171 years with over 225 million business records in their global commercial database.
In the last two decades, investments made in digitalization and transaction automation have enabled organizations to evolve significantly in their ability to capture raw customer behaviour data. Now we find organizations looking to exploit this data through analytics to improve business performance.
The massive increase in computing power, cheap cost of data storage and integrated data supply chains cutting across traditional boundaries are expanding the universe of problems that could be addressed by using analytics and optimization solutions. Key areas where analytics can make a significant impact are: efficient capital deployment, improved speed & accuracy of customer decisions, development of differentiated products based on customers\' needs & risk profiles and reduction in operational costs through enhanced automation in routine decisions; thereby allowing businesses to become more consistent, competitive and customer centric. Therefore, in 2014, we expect the interest and investments in analytics to further grow with the following key trends across demand and supply side.
Evolution of Emerging Markets
In my work with several banks in emerging markets of South Asia, Middle East and Africa, I find that adoption of analytics is less significant compared to those in more mature markets. Availability of sufficient data of acceptable quality and appetite for investments in analytics infrastructure has been the traditional hurdles. A fundamental belief that analytics can produce the right answers, even if such answers are counter intuitive, is sometimes absent in the higher levels of management. This may be because a generation ago, most organizations lived in a data starved environment which resulted in most successful managers of the time relying on experience and intuition to take decisions.
However, over the last few years, there has been a significant change in the mind-set of top level management, even though this is yet to completely permeate the corporate board rooms. This scenario is likely to undergo a change due to the availability of data from past investments in core transactional systems along with significantly reduced costs due to technological innovations. This is likely to encourage more C-suite managers to take the first step and realize the benefits that justify higher investments in greater adoption of analytics for better business outcomes.
Use of Unconventional Data
There is a lot of buzz around \"Big Data\". Market research firm IDC estimates the total revenue for big data to go up to nearly $17 billion by 2015 with venture capital and private equity firms pumping in around $200 million into Big Data and Analytics companies in November 2013 alone. The latest Gartner\'s 2013 hype cycle places Predictive Analytics at a level called \"Plateau of Productivity\". implying widespread usage, placing Big Data at a level defined as \"Peak of Inflated Expectations\". This means that the actual usage and business benefits are still far lower than the hype and expectations.
However, this does not imply that businesses are not exploring the opportunities of innovative use of unconventional data in building new analytical models to address a new set of business problems. Social media has already become a mainstream form of communication, where people provide detailed information about themselves, their preferred products and services, particulars about their daily activities and other important aspects about their personality. This information can help organizations understand consumer preferences and customize their product and service offerings accordingly. However, it is yet to have a significant influence on lending decisions because of lack of reliability, lot of noise and inherent negative bias.
At D&B, we have invested in building sophisticated technology infrastructure that allows us to capture social media inputs on business entities. We are also carrying out research on how our risk models can incorporate these inputs in predicting credit behaviour of business entities by mitigating such biases. Additionally, we also look at new data sources to find appropriate proxies and signals that can add to our conventional sources of data to add more muscle to existing models as well as offer additional analytical interpretations.
Embedding Analytics in all Business Decisions and Processes
For a long time, transactional systems used for business processes automation have been working in isolation. That\'s now changing with rapid advancement in computing power & speed and a new set of tools that put analytics at the core of these systems to support millions of small decisions made by an organization to drive better business outcomes. We expect this trend to further evolve leading to a transition from bureaucratic hierarchies to analytics-driven networks that make every decision optimal. Real time integration between operational systems and analytical models will also facilitate the models to evolve on a dynamic basis to become self-correcting, thereby providing more accurate results.
Analytics on the Go
Rapid technological advances in mobile applications and platforms have enabled a new cost effective channel that is available \"anytime, anywhere\". As mobile devices touch and transform our daily lives in increasingly significant ways, more users will want to gain access to enterprise BI and analytics data using the same devices. It is expected that about a third of BI functionality in 2014 will be consumed by users with mobile devices and they will want the mobile applications to provide a dynamic two way interface rather than mere static reports. This two way interface will also create an opportunity to capture more real time data on customers & suppliers engaged with these users as well as geo-spatial data to further advance the analytics maturity levels in an organization.
Shortage of Analytics Talent
This is the coming of age for the analytics profession as a whole. Given the exciting possibilities that come with the usage of analytics and the declining costs of development and deployment of analytical solutions, organizations are turning to analytics to improve business performance. However, the analytics talent pool with requisite knowledge and skills is still in short supply and the need of the hour is to develop the human capital required to truly exploit these trends. Therefore, we will not only continue to see organisations investing significant resources in training and re-skilling raw talent but also witness an emergence of specialized curriculum in mainstream academic institutions, as well as growth in niche providers imparting analytics skills and courses to capitalise on this opportunity.