Real-Time Innovation: Building a Better Growth Model

The first wave of big data has helped organizations reach a critical point in their transition to becoming more digitally driven. Organizations are collecting and analyzing ever-larger volumes of data, gathered from myriad sources. They’ve begun to streamline processes and systems for managing the data. Some have built big data architectures that span cloud and on-premise systems.

Now comes the hard part.

The next phase of the big data revolution is figuring out new ways to utilize the information to make better decisions, identify new business opportunities and improve the customer experience. We’re really just at the beginning of the data revolution as companies try to figure out how to exploit the rapidly expanding volumes of information they’ve been collecting.

It’s a big challenge: Despite big data investments that IDC predicts will surpass $48 billion globally in 2019, CapGemini last year found that only 13% of companies have “full scale” big data implementations and that only 27% rate them as successful.

“We’re still very much in the early, pioneering days of big data exploration and analytics. We’re just starting to discover the business potential of mining and analyzing this data,” says Joe McKendrick, an independent technology analyst.

The potential spans the entire organization, from back-office operations to customer-facing best practices, across many industries. For example:

  • ThyssenKrupp is gathering data from the cloud-connected elevators it manufactures to improve operations and maintenance. Data from elevators is fed into dynamic predictive models, which continually update datasets, enabling the elevators to actually teach technicians how to fix them – sometimes even before they break down.
  • Royal Caribbean International has connected more than 650 existing and new devices (digital signs, POS terminals and tablets) aboard its line of luxury cruise ships. Harnessing the resulting ocean of data results in powerful insights into what guests want, enabling cruise line staff to create truly personalized guest experiences.
  • The Tacoma Public School District is using predictive analytics tools to provide comprehensive data snapshots of student success indicators. By examining all the data surrounding a student and identifying those who are likely to disengage and ultimately drop out, educators and administrators are empowered to help more students succeed. The approach has helped to improve graduation rates from 55% to 78%.
    video23

As these examples demonstrate, capturing this type of value requires a more sophisticated approach to big data – one that is focused on real-time analytics and predictive modeling, and supported by cloud platforms that are more cost-effective and scalable than legacy data management systems.

A Shifting Landscape

One issue companies are grappling with is the sheer volume of data. Take marketing, for example. Marketing generates a huge amount of data, but only about 28% to 40% is used for business intelligence, according to a 2015 Forrester report.

Marketing is just one of many groups where analytics has not reached its full potential. “Sales might like to keep records on why they lost business and to whom. Manufacturing and operations have so much data about suppliers and products that they do not see the patterns,” says Dan Kusnetzky, chief research officer for the Kusnetzky Group.

That’s where the cloud comes in. The cloud offers a cost-effective way to manage high volumes of data, often with ready-made analytics tools as well as Hadoop implementations for managing and distributing large amounts of data and unearthing hard-to-find relationships between different types
of information.

WASH Multifamily Laundry Systems, for instance, uses sophisticated business intelligence (BI) software to optimize laundry load pricing and maintenance for its 500,000 washers and dryers across 70,000 locations.

With Power BI, a cloud-powered business analytics service, WASH employees visualize and analyze powerful business data anytime, from any device. They’re able to share and discuss insights with colleagues by posting query results in the cloud.

The beauty of Power BI is that employees are creating reports themselves without involving the IT team and sharing those reports with colleagues. This is extremely powerful for us.
Allan Villegas
CFO, WASH

Suppose a financial manager at WASH notices that parts expenses were higher than normal one month. Was it because a certain class of machines began to wear out? Were certain technicians replacing parts more frequently? In minutes, the manager uses Power BI to penetrate heaps of unintelligible data and zero in on the cause: A key supplier raised prices.

The above example highlights a vital component of a successful big data program: user-friendly data visualization tools.

“The right visual front ends make data trends understandable to anyone who views them,” says McKendrick. “With data visualization, end users can quickly assess trends and spot outliers – often with a single, split-second glance.”

Allan Villegas, WASH’s chief financial officer, concurs.

“In the past, our financial manager would have had to extract data, send it to the IT department and wait weeks for a custom report,” says Villegas. “The beauty of Power BI is that employees are creating reports themselves without involving the IT team and sharing those reports with colleagues. This is extremely powerful for us.”

Cloud Opens New Doors for Big Data

The prerequisite infrastructure is falling into place as cloud deployments increase. The 2015 IDG Enterprise Cloud Computing Study found that nearly three-quarters of organizations have moved at least one application or a portion of their IT infrastructure to the cloud. Cloud deployments currently make up 45% of the average IT environment but are expected to comprise 56% of the IT ecosystem in the next 18 months.

cloud-models
Source: 2015 IDG Enterprise Cloud Computing Survey

While the first wave of cloud deployments has focused largely on cost efficiencies and replacing legacy technology, organizations are starting to turn their attention to more business-focused objectives such as enabling innovation, new product development and expanding revenue opportunities. Capturing those business benefits will likely require a heavy dose of data and analytics. Predictive analytics can help identify new markets, create up- and cross-sell opportunities for existing customers, increase customer loyalty and identify problem areas before customers defect. They can also change the face of an entire industry. Take healthcare as an example.

“The current healthcare system asks a patient to come to us when it’s convenient for us,” says Dr. Ethan Berke, medical director for Clinical Design and Innovation at Dartmouth-Hitchcock Medical Center in Lebanon, N.H. “We asked, ‘How can we meet a patient where they are? Instead of being a reactive health system, how can we become a proactive health system?’”

The answer is ImagineCare, a system Dartmouth-Hitchcock is piloting to 6,000 patients that embraces the idea of the medical system going to patients. Patient devices such as blood pressure cuffs, oximeters and activity trackers feed data into the Microsoft Azure cloud. That data is massaged by the Cortana Analytics Suite to determine if any patient thresholds have been exceeded. If so, a nurse reaches out to the patient by phone, video chat or secure text.

We’re truly looking at the whole health of the individual. It’s a completely new way to deliver care.
Nathan Larson
Director of Remote Medical Sensing,
Dartmouth-Hitchcock

Each patient’s personal care plan is continually updated in a secure, HIPAA-compliant system, where machine-learning algorithms examine not only the person’s own data but also near-real-time data from similar people based on location, age, disease, biometrics and more. What results is a dynamic and contextual care plan that is infinitely more customized — and definitively more accurate — than traditional plans.

In the future, imagine integrating personal data such as caloric intake, exercise activities and heart rate from wearable devices to address lifestyle issues.

Where ImagineCare breaks into uncharted territory in healthcare is with its ability to track mental and emotional health. Using perceptual intelligence capabilities, the system can detect a person’s emotional state by monitoring Twitter feeds and other social media to perform sentiment analysis, looking for troublesome trends. It can also perform speech and tone analysis during interactions with ImagineCare nurses — a boon for early intervention. A mobile app that invites timely mood check-ins completes the picture, vastly increasing the chances of catching and treating depression before it becomes a life-changing issue.

“We’re truly looking at the whole health of the individual,” says Nathan Larson, director of Remote Medical Sensing at Dartmouth-Hitchcock. “It’s a completely new way to deliver care.”

And a completely new way for any organization to think about big data in the enterprise.