Almost every conceivable enterprise from large industrial conglomerates, hospitals, humanitarian organizations, and professional sports leagues to municipal police forces, consumer products companies and weather forecasters are capturing trillions of bytes of information and using it as a catalyst to transform their organizations by developing better products and services, strengthening customer relationships, enhancing operational efficiency and even improving the lives of the less fortunate.
The phenomenon of big data sprang from the emergence of global Internet companies whose business models were based on the collection and analysis of large amounts of data. As the tools to collect data became more widespread and efficient, (millions of networked sensors are being embedded in industrial robots, mobile devices, drones, cameras, microphones, automobiles, jet engines, RFID readers, etc.) the need for hardware and software to communicate and analyze it exploded.
“I think Jay is in the import and export business as his cards say, but he finally found that the second most valuable commodity today is information."
"The most valuable?"
"People with information," I suggested.”
The Ipcress File
In Len Deighton’s 1960’s best seller, a gritty, downbeat, counterespionage thriller, he underscores how information was just as important to his fictional spy of the 60’s as it is to today’s managers- it allows them to measure and therefore understand more about their enterprises and translate that knowledge into vastly improved decision making and results.
Those results are meaningful. In a Harvard Business Review survey, one correlation was readily apparent: the more companies defined themselves as data-driven, the better they performed in terms of financial and operational results. Companies that extensively employ big data-driven decision-making were 5% more productive and 6% more profitable than their competitors.
So just what is big data? Most simply, it is defined as a volume, variety, and velocity of data that is too large and complex for processing by traditional database management tools. Its size and complexity requires the application of high-level computing power in the form of machine to machine (M2M) connectivity, learning and artificial intelligence in order to collect, categorize, analyze, store and extract actionable intelligence.
The phenomenon of big data sprang from the emergence of global Internet companies whose business models were based on the collection and analysis of large amounts of data. As the tools to collect data became more widespread and efficient, (millions of networked sensors are being embedded in industrial robots, mobile devices, drones, cameras, microphones, automobiles, jet engines, RFID readers, etc.) the need for hardware and software to communicate and analyze it exploded. And there’s a lot of it. IDC, a technology research firm, estimates that the amount of data generated by all these devices is doubling every two years. Big data is now an intricate and indispensible part of the global economy.
Now almost every conceivable enterprise from large industrial conglomerates, hospitals, humanitarian organizations and professional sports leagues to municipal police forces, consumer products companies and weather forecasters are capturing trillions of bytes of information and using it as a catalyst to transform their organizations by developing better products and services, strengthening customer relationships, enhancing operational efficiency and even improving the lives of the less fortunate.
General Electric has established a global software center in San Ramon, California devoted to achieving their vision of the industrial Internet. Using their software platform Predix, the goal is to make use of the gigabytes of digital data coming from sensors embedded in GE’s jet engines, turbines (a typical GE gas turbine generates 500 gigabytes of data daily), utility power equipment, trains and hospital MRI devices for the benefit of their customers.
For example, GE is designing its next version of jet engines to record information on engine performance for every flight. The fuel efficiency gains from such information could result in as much $2 billion savings for their airline clients over the next 15 years.
This initiative will also help GE’s other customers in their power, healthcare, rail and oil and gas businesses to spot potential maintenance problems and make improvements could eventually translate into $300 billion of savings by 2030.
Infinity Insurance processes over 25,000 claims in a typical month. Being able to determine which ones are fraudulent is an enormous challenge not only for them, but also for the industry. The National Insurance Crime Bureau predicts fraud is a $20 billion risk to insurance companies each year. Infinity wanted a way to pinpoint questionable claims faster and speed processing for legitimate losses.
Using predictive analytics and text mining the thousands of claim reports the company had on file, Infinity was able to provide a more accurate way to identify fraud. Additionally, they were able to use that data to re-engineer the entire claims processing workflow. The result was less reliance on external adjusters, lower costs and faster claims processing.
Consequently, the time it took Infinity to detect fraudulent claims went from 14 days to 24 hours, which improved customer satisfaction dramatically. Their return on investment for the first 90 days alone was over 400%.
Barclays Bank, following a major strategic review, identified five key values as critical to its future: respect, integrity, service, excellence and stewardship. In order to insure that these values were embedded into everything they do, they looked for a method of generating ideas from their entire workforce. The solution was Barclays Value Jam, an intuitive web portal that received over 35,000 posts in an open and honest company-wide conversation. The initiative yielded a treasure trove of ideas, solutions and actionable outcomes that would make that goal a reality and have a measurable impact on Barclay’s worldwide franchise.
To deal with the immense volumes of data generated by the event, Barclays used advanced text mining and data analytics to create heat maps and word clouds in order to surface out keywords, themes and overall topics for further exploration. From this, Barclay’s established16 collaborative workshops that were used to understand, confirm and catalogue the results. In less than two weeks, they identified the highest priority topics and less than two months after that Barclays had a list of the 22 most important themes for immediate focus.
UPMC is a $9 billion global health enterprise with more than 54,000 employees headquartered in Pittsburgh, Pa. At their newly finished Center for Innovative Science, they are employing advanced analytics and genetic sequencing to focus research on personalized medicine, cancer biology and the aging process.
Personalized medicine is the ability to develop an individualized treatment plan for every single patient. For diseases like cancer, UPMC will be able to identify the unique genetic and environmental factors that determine the susceptibility of each individual and design the best course and type of treatment.
By combining clinical, genomic, financial and administrative data from what is now 200 discrete systems, UPMC believes this will result in healthier patients, better-informed doctors and nurses, advance research, reduce over-diagnosis and unnecessary treatments all while lowering costs.
The London Metropolitan Police Service employs 31,000 officers, 13,000 police staff and 2,600 Police Community Support Officers (PCSOs) and is responsible for covering 32 boroughs encompassing an area of 620 square miles. Their goal is to protect and serve their 7.2 million citizens and to make London the safest major city in the world.
As in any major city, police resources are limited and need to be used as effectively as possible by targeting them toward the highest risk individuals. In 2013, the Metropolitan Police Service embarked on an initiative to use data analytics to fight gang crime in London. Their twenty-week pilot program merged and assessed a wide range of crime reporting and criminal intelligence data ranging from previous crimes to social media activity from the period 2009 to 2012.
Predictive analytics were applied to calculate the likelihood of gang members committing violent crimes in 2013. That data was then compared to actual known gang crime activity to see how accurately it could be predicted. The software used can also forecast other crimes such as burglary merely by tweaking the algorithm.
The pilot program was completed in October of 2014 and the results have not yet been made public.
The National Football League is making use of big data to improve fan experience and to help their teams make better player acquisition and game strategy decisions.
Fantasy football fans are devoted to the game more than other fans. Besides following their favorite team, they compete with other enthusiasts by picking a group of individual players in hopes that their performances during the season will exceed that of their fellow fantasy competitors. In an effort to leverage the vast interest in fantasy football, the NFL is tapping their immense data reserves and providing real-time recommendations on over 100 statistics from each player. Partly as a result of the NFL’s big data initiative, participation in fantasy football is at 33 million, up considerably from last year’s 27 million.
As well, the League is also making available to their 32 teams their massive statistical pool in order to enhance scouting programs, make more accurate player acquisition decisions and develop more effective weekly offensive and defensive game-plans. NFL Vision Platform provides video clips of each recorded statistic by player including playing surface, stadium, weather conditions, etc. Additionally, college teams are supplementing this information by providing video from the player’s college performances.
Several NFL teams are analyzing data on everything from opponents’ defensive alignments to injury likelihood. The Jacksonville Jaguars for example, used NFL Vision Platform to thoroughly evaluate offensive tackles available in the draft and determine which player to select with their second pick.
The Coca-Cola Company generates petabytes of data from both online and offline customers, consumers, retail transaction scan and merchandising data, social media, loyalty cards, manufacturing, and sales and shipment data from their bottling partners.
One of their key business objectives is to use all of this information to analyze consumer trends and make better, faster decisions about marketing and distribution. Eventually this will help their retail partners become more efficient and assist them in driving sales, improving the experience for shoppers and ensuring that Coca-Cola delivers the right products to the right stores at the right time to meet shoppers’ demand patterns.
For instance, they employ high-level data analytics to decide the optimal blend for orange juice in order to create a consistent taste and pulp content throughout the year despite a peak-growing season of just 90 days. They have developed an algorithm, that combines data such as satellite imagery, weather, expected crop yields, cost pressures, regional consumer preferences, data about the hundreds of different flavors that make up an orange, as well as acidity or sweetness levels to tell them how to blend the orange juice to create a consistent taste. It is one of the most complex applications of business analytics requiring the evaluation of up to 1 quintillion variables.
WSI, a division of the Weather Company, specializes in providing their clients the world’s most accurate weather forecast. Their customers include top companies in the media, aviation, energy, insurance and utility industries and they depend on WSI to provide them with weather related information that is critical to their businesses.
The world’s airlines rely on accurate weather forecasts to keep their operations running smoothly and efficiently. In order to meet their aviation customer’s needs, WSI offers Fusion, the world’s most powerful forecast model consisting of the most wide-ranging and innovative collection of aviation weather products for pilots and operations personnel on the market today.
WSI Fusion is an operations management solution that analyzes an immense amount of data including public and proprietary weather information, airspace constraints, and flight information and derives a comprehensive operational picture in order to enhance decision making. WSI Fusion™ helps dispatchers, operations personnel, and managers get out in front of changing conditions, alleviate the impact of disruptive events, and stay on plan, on time, and on budget
As part of their analytics suite, the Airport Congestion Index is a unique data-driven program that can predict taxi delays and deicing impacts. By providing timely forecasts, operations personnel can determine when increases in fuel consumption, crew time or deicing fluid may be required- minimizing delays, improving passenger safety and maximizing efficiency.
The United Nations Global Pulse initiative is an attempt to facilitate a network of innovation labs with the goal safely harnessing big data for the good of the planet. The U.N. is partnering with experts from its own agencies, governments, academia and the private sector in an effort to achieve a critical mass of global best practices, lower barriers to implementation and reinforce the big data innovation network.
Through their Pulse labs in New York, Jakarta and Kampala, Uganda they are endeavoring to use the vast digital trails left behind by mobile devices, on-line shopping, money transfers, Internet searches and social network activities to analyze population behavior, assist in humanitarian efforts, identify the most promising areas for sustainable development, track emerging market trends and potential crises and get real-time feedback on the efficacy of their policy responses.
One effort studied the correlation between mobile phone airtime credit purchases and the consumption of several food items, such as vitamin-rich vegetables, meat or cereals in East Africa. The findings demonstrated that airtime credit purchases were highly correlated to marketplace food expenditures and could be used to help estimate levels of poverty in areas where standard household surveys are not always feasible.
As with all technological innovations that have impelled humanity to greater levels of productivity and comfort, the ultimate measure of success of big data is how we exploit it. Big data is a tool and like its predecessors the wheel, the bow and arrow, the printing press, electricity or the personal computer, requires man’s intuition, creativity, imagination, morality, artistry, and emotion to fully realize its potential. The future narrative of big data and whether it will be as profoundly impactful as the Internet lies not in its technological wizardry but in our transformative wisdom. It’s not information but people with information that is the most valuable commodity.
Arthur C. Clarke, futurist and renowned science fiction author cautions us on the essential nature of the man/machine symbiosis - “Before you become too entranced with gorgeous gadgets and mesmerizing video displays, let me remind you that…. information is not knowledge, knowledge is not wisdom, and wisdom is not foresight. Each grows out of the other, and we need them all.”