Big Data Requires Strong Analytics Capabilities

Global Banking Industry Marketing, Big Data, IBM

This is our fifth post in a series of seven presenting the findings from the IBM Institute for Business Value and University of Oxford’s big data study, “Analytics: the real world use of big data in financial services.”

As part of this recently published global research study, my colleagues David Turner, Michael Schroeck and Rebecca Shockley found that big data itself does not create value until it is put to use to address important business challenges. This requires access to more and different kinds of data, as well as strong analytics capabilities that include not only the tools, but the requisite skills to use them.

By examining the banking and financial markets companies that are engaged in big data activities, the study reveals that those companies start with a strong core of analytics capabilities designed to address structured data, such as basic queries, predictive modeling, optimization and simulations. However, they lag behind their cross-industry counterparts in core capabilities of text analytics and data visualization (see figure below).



The need for more advanced data visualization and analytics capabilities increases with the introduction of big data. Datasets are often too large for business or data analysts to view and analyze with traditional reporting and data mining tools. In the 2012 Big Data @ Work Survey conducted by The IBM Institute for Business Value and the Saïd Business School at the University of Oxford, banking and financial markets respondents said that only three out of five active big data efforts utilize data visualization capabilities.

Big data also creates the need to analyze multiple data types, and this is where banking and financial markets firms lag significantly behind their peers in other industries. In fewer than 20 percent of the active big data efforts, banking and financial markets respondents use advanced capabilities designed to analyze text in its natural state, such as the transcripts of call center conversations. These analytics include the ability to interpret and understand the nuances of language, such as sentiment, slang and intentions, and are often used to bolster efforts to understand behavior and preferences and improve the overall customer experience.

Fewer than one in 10 active big data efforts in banking and financial markets report having the capabilities to analyze even more complex types of unstructured data, including geospatial location data (7 percent), voice data (10 percent), video (7 percent) or streaming data (6 percent). While the hardware and software in these areas are maturing, the skills are in short supply. Additionally, banks are still struggling to monetize these capabilities.

The current pattern of big data adoption highlights banking and financial markets companies’ hesitation, but confirms interest too. In order to better understand the big data landscape, we asked respondents to describe the level of big data activities in their organizations today. The results suggest four main stages of big data adoption and progression along a continuum that we have identified as Educate, Explore, Engage and Execute (see figure below).


  • Educate – Building a base of knowledge: 26 percent of banking and financial markets respondents
  • Explore – Defining the business case and roadmap: 47 percent of banking and financial markets respondents
  • Engage – Embracing big data: 23 percent of banking and financial markets respondents
  • Execute – Implementing big data at scale: 3 percent of banking and financial markets respondents

At each adoption stage, the most significant obstacle data efforts reported by banking and financial markets firms is the gap between the need and the ability to articulate measurable business value. Executives must understand the potential or realized business value from big data strategies, pilots and implementations. Organizations must be vigilant in articulating the value, forecasted based on detailed analysis when needed and tied to pilot results where possible, for executives to commit to the investment in time, money and human resources necessary to progress their big data initiatives.

In our next blog, we will review some initial recommendations for companies that want to extract more value from big data.

To learn more