Big data trends: The top eight analytics lessons for business
I have been working in IT for more than 14 years, yet I’ve never stopped being amazed by how quickly technology evolves to meet business needs. The evolution of technical computing has progressed from the first computers—which filled entire rooms—to mainframes, then to personal computers, then to the web and software as a service (SaaS) applications, then to mobile devices and now to cloud computing and the Internet of Things.
Thanks to that long progression, big data is the world’s newest natural resource. The value of big data is self-evident, and the need for powerful analytics capable of delivering that value is pressing in the business environment. Business leaders no longer question whether value is to be found in data; instead, they wish only to learn how to extract that value to understand their customers and meet critical business needs.
This desire has given birth to a generation of enterprises that are data-rich and analytically driven, eagerly following trends in big data and analytics. Let’s take a closer look as I provide some use cases demonstrating how IBM is helping clients find innovative big data solutions.
1. Datafication-led innovation
Data is the new basis of competitive advantage. Enterprises that use data and sophisticated analytics turn insight into innovation, creating efficient new business processes, informing strategic decision making and outpacing their peers on a variety of fronts.
IBM Case Study: Fannie Mae
Fannie Mae provides a great example of achieving competitive gains through innovation enabled by data analytics. The company began its Vega Analytics initiative with the goal of developing a comprehensive set of analytics tools for collateral valuation and risk management. Fannie Mae improved its loan origination and securitization processes by using the advanced edition of the IBM Platform Symphony solution as its big data analytics platform, together with IBM Software Defined Infrastructure. Fannie Mae enhanced its financial risk and fraud analytics, and the project won “best analytics initiative” of 2014 in the American Financial Technology Awards.
2. Sophisticated analytics for rich media
Much of produced data is useless without applying appropriate analytics to it. Where does opportunity lie? According to the International Data Corporation (IDC), rich media (video, audio, images) analytics will at least triple in 2015 to emerge as a key driver for big data and analytics technology investment. And such data requires sophisticated analytics tools. Indeed, consider e-commerce–based image search: accurate, relevant image search analysis that doesn't require human tagging or intervention is a significant opportunity in the market. We can expect similar smart analytics capabilities to offer similar opportunities.
IBM Case Study: Television Broadcasts
Hong Kong’s first wireless commercial television station, Television Broadcasts Ltd. (TVB), implemented social media analytics to help boost its ratings, mining more than three decades’ worth of program ratings data to understand trends in media consumption. TVB worked with IBM Business Partner Big Data Architect to develop a business intelligence and social media analytics solution drawing insight from raw, uncensored viewer reactions communicated through social media. The solution provides a solid understanding of why some shows succeed and others fail—whether timeslot, subject matter or star power—allowing the network to adjust its programming to boost viewership and demonstrate its relevance to advertisers.
3. Predictive analytics driving efficiency
Applications featuring predictive capabilities are picking up speed. Predictive analytics enhances value by boosting effectiveness, providing measurability of the application itself, recognizing the value of the data scientist and maintaining a dynamically adaptive infrastructure. For these reasons, predictive analytics capabilities are becoming an integral component of analytics tools.
IBM Case Study: PinnacleHealth
With healthcare providers under pressure to reduce readmissions for patients who have chronic health conditions, the limited resources available to providers make understanding which patients require the most attention vitally important. PinnacleHealth needed to build a predictive model evaluating risk of readmission for patients who had chronic obstructive pulmonary disease (COPD), enabling effective intervention at the point of care. By using IBM Cognos Business Intelligence to help predict COPD readmissions with 85 percent accuracy, PinnacleHealth enhanced patient outcomes and cut costs. And its success in doing so is an encouraging sign for healthcare providers hoping to use similar models to assess other chronic conditions.
4. Big data in the cloud
Over the next five years, IDC predicts, spending on cloud-based big data analytics solutions will grow three times more quickly than spending on on-premises solutions—and hybrid deployments will become a must-have. Moreover, says IDC, with data sources located both in and out of the cloud, business-level metadata repositories will be used to relate data. Organizations should evaluate offerings from public cloud providers to seek help overcoming challenges associated with big data management, including the following:
- Security and privacy policies and regulations affecting deployment options
- Data movement and integration requirements for supporting hybrid cloud environments
- Building a business glossary and managing map data to prevent overwhelming data
- Building a cloud metadata repository (containing business terms, IT assets, data definitions and logical data models) that points to physical data elements
5. Cognitive computing
Cognitive computing is a game-changing technology that uses natural language processing and machine learning to help humans and machines interact naturally and to augment human expertise. Personalization applications using cognitive computing will help consumers shop for clothes, choose a bottle of wine or even create a new recipe. And IBM Watson is leading the charge.
6. Big money for big data
Increasingly, organizations are monetizing their data, whether by selling it or by providing value-added content. According to IDC, 70 percent of large organizations already purchase external data, and 100 percent are expected to do so by 2019. Accordingly, organizations must understand what their potential customers value and must become proficient at packaging data and value-added content products, experimenting to find the “right” mix of data and combining content analytics with structured data, delivered through dashboards, to help create value for external parties interacting with the analysis.
7. Real-time analytics and the Internet of Things
The Internet of Things (IoT) is expected to grow at a five-year CAGR of 30 percent and, in its role as a business driver, to lead many organizations to their first use of streaming analytics. Indeed, the explosion of data coming from the Internet of Things will accelerate real-time and streaming analytics, requiring data scientists and subject matter experts to sift through data in search of repeatable patterns that can be developed into event processing models. Event processing can then process incoming events, correlating them with relevant models and detecting in real time conditions requiring response. Moreover, event processing is an integral part of systems and applications that operationalize big data, for doing so involves continuous processing and thus requires response times as near to real time as possible.
IBM Case Study: Memorial Hermann Health System
Memorial Hermann Health System teamed up with IBM to change its data storage methods, hoping to identify illnesses early by shortening medical record access times—keenly aware that doctors need rapid, reliable access to medical records and ready insight into changing health indicators if they are to make ideal decisions. The organization deployed an extremely high-performance flash storage system providing rapid access to patient records and supporting real-time analysis of medical data. In doing so, Memorial Hermann Health System cut average response times for its medical records database by more than 99 percent, enhancing treatment decisions.
8. Increased investments in skills
Many organizations want to combine business knowledge and analytics but have difficulty finding individuals who are skilled enough to do so. Leading companies in particular feel this talent gap keenly, for as they move to broaden skills across the enterprise, the need for combined skills becomes ever more apparent. Indeed, combined skills are of critical importance in speed-driven organizations, for such skills speed the translation of insights into actions through deep knowledge of the business drivers—and the data related to them—that are likely to affect performance.
Setting out on your analytics journey
As you begin your analytics journey, start with your people. Then build a culture that infuses analytics everywhere. Seek out helpful business use cases, and apply analytics in ways that improve your core competitiveness, building against a master plan and investing in capabilities deployable both on premises and in the cloud.
Can you identify other noteworthy trends in big data and analytics? Continue the conversation with me on Twitter, then deepen and broaden your experience of the big data and analytics platform by exploring IBM Analytics. Even better, register to attend IBM Insight 2015, October 26–29, 2015, in Las Vegas, Nevada.