Commentary - Analytics are increasingly being classified as a “must have” for organizations, but we’ve merely scratched the surface of their potential.
Just look at the machines at the neonatal ward at the University of Ontario in Canada, which don’t buzz or beep any differently than those in other hospitals. What sets them apart, and what’s transformed the center’s infant care and survival rates, is a system in the background that automatically analyzes vital statistics in real time and pinpoints patterns so that the staff can head off problems before they escalate.
In Washington DC, DC Water, the local water and sewage authority, uses real-time data and analytics to uncover patterns related to weather conditions, water use and demographics, allowing the utility to act before sewer backups, equipment failure, and pipe leaks occur, while at the same time, saving money and water.
Companies and government agencies previously used data to understand what has happened in the past. Now they want to learn what is happening now. They are increasingly applying new predictive analytics technologies that pick up on patterns and statistical correlations within massive streams of data, allowing them to anticipate and plan for events before they happen. This is allowing them shift from being reactionary to preemptive.
But that is just the tip of the iceberg.
The time is ripe for predictive analytics that go beyond customer data. Business leaders crave more insight into the day-to-day operations of their own organization.
Just look at today’s typical IT systems. They are a patchwork of different services all within an increasingly smarter infrastructure: cloud computing, enterprise mobility, IT infused assets and dynamic Web accessible data. As a result, today's IT leaders are asking themselves -- where does our business infrastructure begin and end these days? How do I best optimize it to drive desired business outcomes when its appears to be without limits?
In an environment where the demands of the IT are ever-changing, it is crucial for the business to gain visibility, control and automation of their infrastructure. Thousands of IT events happen every day--ranging from service desk needs to network hiccups-- so having greater insight into the monitoring and performance will drive optimal business operations.
Organizations simply can’t afford to throw more people at this problem. Instead they have to get smarter by turning predictive analytics know-how inward to provide insight to a firm’s own internal functions – the data center, operations, business processes and software delivery and development life cycle.
Predictive analytics can review and visualize all the relationships and performance issues in a system to identify and predict trends or patterns on how efficiently the business is running. In fact, by looking at streaming data rather than only historical data, organizations can perform real-time analysis to predict potential application and service degradation or outages before they happen.
This can significantly reduce network congestion, prioritize urgent service requests, or dispatch data center engineers onsite before critical business services behave erratically. More importantly, predictive analytics can move the management of IT from a reactive to preemptive process to make more intelligent, automated business decisions.
For example, Swiss Federal Railways collects data from more than 50,000 fixed and 20,000 mobile connections. The railways uses analytic software to perform diagnostics to identify irregularities that may signal a potential outage of services and automatically takes steps to resolve them, such as scheduling and alerting Swiss rail staff to perform maintenance. Analytics helps to recognize and repair more than 50 percent of network issues that could cause train delays before they occur, leading to an annual savings that's equivalent to US$ 2.3 million.
If there’s ever a time when we need predictive analytics within IT, it’s now. For the teams responsible for managing the complexity of IT, the pressure to prevent downtime is relentless. The amount of data, the seemingly limitless places from which organizations are collecting information, and the velocity with which we’re gathering it means that firms are being surrounded with more data than they can make sense of.
Organizations can no longer afford to analyze data in a "rear-view mirror" only. Predictive analytics allows them to analyze the trends and patterns in both historical and current data in order to predict what will happen in the future. The time to act is now, because predictive analytics may provide a competitive advantage right now, but soon will be the best way of doing business.
biography Scott Hebner is vice president of software for Tivoli, IBM