Big Data & Insights: Oil for the Digital Age
By Radhika Venkatraman, SVP - CIO, Network & Technology, Verizon
Sir Arthur Conan Doyle wrote “It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.” That’s as true today, in the era of big data, as it was it was in 1891.
Until recently, Verizon and most large organizations were drowning in data with limited ability to transform information into actionable insights. Decisions were driven by experiences and intuition, twisting fact to suit theories, often in response to a specific issue. Today, with the help of new data mining techniques and methods of analysis, we’re leveraging the power of Big Data to provide more efficient, predictive and, perhaps most importantly, personalized customer experiences.
Each and every day, Verizon network devices, front line employee tools and digital customer interactions are generating terabytes upon terabytes of service and diagnostic information. The challenge for Verizon was how to use this precious resource and draw valuable insights that are not only actionable but can be processed to provide a continuous feedback loop.
Verizon embarked on the journey of building a robust data science practice. The first step was to build a low-cost scalable infrastructure with the ability to process real-time streams of data. The next step was to install analytics and visualization toolkits to help technologists and business leaders work through the process of ideation, experimentation, and investigation. Equally important to the technology, the third step was to create a data-driven culture focused on using that data to improve customer experience.
“As digital becomes the new normal and the explosion of connected technology continues, dependence on data science and algorithmic prediction will increase”
The technology teams partnered closely with the business and operations teams and applied Six Sigma principles. These cross-functional groups filtered hundreds of opportunities to create the right pipeline of problem sets to examine. By putting our customers at the center of the journey they were able to select and prioritize data sets with the highest potential for positive impact on customer experience. As the infrastructure and culture matured, the teams began to set up sustainable processes based on feedback loops and continuous improvements.
The examples illustrated here are all built on the premise that we could leverage the power of data and find better and more efficient ways to take care of our customer’s unmet and unrequested needs.
Can we prevent issues before they occur?
“If a tree falls in the forest and no one is around to hear it, does it make a sound?” The winter of 2014 brought record snowfall in much of the Northeast, toppling trees onto overhead phone lines. It’s important to note that sometimes a phone line doesn’t snap right away. Instead, a tree may stay precariously waiting for a final gust of wind or bit of ice before taking down the phone line. As the fallen tree sits pressing on the phone lines, data transmission is impacted. Indicators, such as optical power, jitter, delay and latency, show abnormal values during the ‘tree event’. Our real-time analytics engine correlates these data streams with normal day patterns to predict a developing situation and proactively alert operations, who can then address the situation before customers experience an outage.
Can we help our customers solve their own issues?
One of our large enterprise banking customers was able to solve a problem even before it happened. The analytics streams detected abnormal data patterns coming from the customer’s equipment. The anomalous behavior served as a red flag, signaling a need for the customer to change a configuration parameter. Verizon contacted the customer and the timely change prevented a large-scale disruption of the bank’s voice communication service.
Can we better serve our customers in the field?
Long gone are the days when customers will accept all day appointment windows. But, how can we shorten the window down to one hour and guarantee never to miss it? The complexity was tamed by correlating information about a field technician’s work location, customer’s job status, job priority, and historical completion trends, and thus predicting an accurate one-hour technician arrival window on the morning of the customer’s appointment.
As digital becomes the new normal and the explosion of connected technology continues, dependence on data science and algorithmic prediction will increase. At Verizon, we have already taken leaps to transform the business model from ‘data as advisory’ to ‘data as actionable’ and are now approaching “data as actionable in real-time.” Nearly every industry is in the midst of disruptive change and we at Verizon are excited to not only be enabling this digital disruption but also carrying the torch to lead through this change.
For every case study we explore and enable, there are hundreds more waiting. For every segment of data collated, there are new devices coming online daily with even more insights to share. Every day we’re excited about the new possibilities of leveraging the power of Big Data to tackle the most complex challenges.