Subscribe / Unsubscribe Enewsletters | Login | Register

Pencil Banner

Predictive analytics help solve intractable problems

Mark Jaffe, CEO of Prelert | Aug. 22, 2013
Predictive analytics can discover and provide the information to address application performance issues before they're even noticed

The beauty of the best predictive analytics solutions is they don't require the lengthy, complex installation and configuration process of most of today's application and network management tools. Nor do they rely on the user to make judgments about which KPIs and thresholds to monitor.

Instead, they use their complex machine learning algorithms to learn about your systems and all the underlying relationships and event chains themselves. Then further algorithms predict the probability of various events and alert you when those that are highly improbable happen.

Unlike KPIs, which may alert you a thousand times to some minor abnormality in one system, predictive analytics tools look at abnormalities across different systems that typical thresholds and alerts often miss. And because they learn about systems over time, they can continually self-adjust to legitimate changes in your IT environment. With typical management tools today, the user has to make those changes.

Aside from predicting and addressing performance issues, predictive analytics solutions are great at providing insight into system relationships and problems you never even knew existed. They've been called the "Donald Rumsfeld of application performance management" because they discover the "unknown unknown" behaviors in your application infrastructure.

A recent survey by TRAC research found 60% of IT organizations report a success rate of less than half when it comes to preventing performance issues that have an impact on end users. It also found that, on average, 46.2 hours are spent each month on "war room" scenarios.

Many of these organizations already have application performance management and numerous other management tools deployed. The addition of a predictive analytics solution can raise that success rate substantially and reduce or even eliminate those time-consuming, resource-intensive war room scenarios. And in doing so, they can improve customer satisfaction and allow IT to spend less time troubleshooting performance issues and more time on strategic initiatives that add value to the business and improve its competitive position.

Jaffe is a serial software entrepreneur who has been instrumental in the success of numerous software companies.


Previous Page  1  2 

Sign up for Computerworld eNewsletters.