Big data innovations turned into big data implementations in my next session of the day. The session began with a recap of the findings from the just-released IBM/Oxford study, “Analytics: The Real-World use of Big Data," followed by a panel discussion featuring four IBM clients with a long track record implementing big data projects.
Study findings were presented by Michael Schroeck, VP & Global Leader, Information Management Foundation, Business Analytics & Optimization, IBM Global Business Services. Schroeck is also one of the report authors. Among the findings:
Over the past two years, the percentage of organizations reporting competitive advantage from analytics has jumped from 37 percent to 63 percent.
Yet, organizations are struggling to leverage the four “Vs” of big data. Specifically:
- volume – from terabytes to petabytes
- variety – structured, unstructured, even semi-structured
- velocity – data in motion, data in streams
- veracity – data uncertainty, separating signal from noise.
Organizations report different levels of maturity with their big data projects:
24 percent are increasing their awareness of big data potential.
47 percent are developing plans and blueprints, teams and roadmaps
28 percent are implementing proofs of concept (POCs), pilots and enterprise solutions. This last group is “very committed and moving quickly.”
The survey highlighted five key findings on how organizations are moving forward:
- Customer analytics are driving the majority of projects.
- Big data success depends on a scalable and extensible information foundation.
Organizations are focusing their Initial big data efforts on gaining insights from existing and new internal sources
- Big data success needs strong analytics
People skills not keeping pace. Organizations are suffering from an analyics skills gap, which is one of the biggest inhibitors of big data progression
Big data projects will not move forward without a strong business case. This finding cuts across industries
To moderate the panel discussion that followed, Schroeck handed the microsphone to his IBM counterpart Sharon Hodgson, Servicve Line leader for North America, Business Analytics & Optimization, IBM Global Business Services. Here are some highlights from their discussion:
- Big data can help big organizations restore the personal touch - at scale. An indivdual insurance agent can have up to three thousand customers. Yet through analytics and natural language processing, the company can discover, optimize and disseminate best practices for customer engagement that improve the personalized dimension of their relationships for all its reps and through all its channels.
- Big data may be big, but it's not complete, at least not yet. Some organizations are missing important data sets to complete their customer profiles. Also, some data sets may be too old to be helpful. Organizations need to take a step back and carefully think about how to get the data they need. “We need to treat data with more respect,” said one panelist. In addition, not every customer is willing to share every piece of information. Organizations need to determine the data sets they need for their models to provide new products or improve existing ones so that all customers may benefit and justify their rationale for asking customers to share.
- The skills shortage is real. Organizations are seeking people who possess not only strong analytics and data skills, but product and industry knowledge as well. Old-school problem-solving is also desirable. Such individuals are in extremely short supply.
- Business leaders must take the lead. Companies will go further with their big data projects and receive more value from them with a business executive at the helm. Also, starting with a clearly articulated desired business goal is key.
- Governments and public leaders must lead the discussion on big data use. The boundaries, roles and responsibilities of organizations pursuing value from big data are fuzzy at best. A discussion on the ethical use of big data needs to happen now.