The Secret to Good Problem Management
Any strong problem management program requires expert knowledge — but the way you leverage that expertise is how you make your money.
As fast as IT technology advances, one lingering question has always been: “Can IT be used to solve its own problems?” No IT infrastructure is perfect, which means that a reliable IT program is earned through consistent and effective troubleshooting to pin down root problems and work out solutions — all to avoid the dreaded crash or slowdown.
However, strategies to find root causes are nearly as numerous as the problems themselves, and many IT professionals disagree on best practice. Writing for Computer Weekly, Michael Hall voices a common concern: many problem management programs lack trained problem managers who can offer structured guidance, causing them to completely lose coherence.
But while it’s important to have solid staff, experienced managers are costly. And given the staggering complexity of IT systems today, they inevitably suffer from a lack of knowledge and time — it will soon be virtually impossible to cover every base.
But virtually, root cause analysis is effective and affordable. Algorithms can process IT systems’ massive data output and pinpoint an exact problem location, in addition to tracking the ongoing health and relative risk of systems. As a bonus, when multiple algorithms are employed, you can augment both your perspective on problems and the veracity of solutions.
This is still anchoring your program in expertise — it’s just built into the software.
In general, it’s these are tough times for knowledge workers like problem managers. Everyone from the International Institute for Analytics to Gartner has predicted that big data processing and automated analytics software will become fundamental to business, subsuming the roles of many knowledge workers along the way. At the very least, these developments will dramatically change the way they operate. It all boils down to efficiency.
No matter the skill of a professional, they face competition that operates on scales of a different magnitude — where a problem manager directs a team to individually parse data sets, analytic software can crunch an entire IT data database in near simultaneity. Machine learning techniques automatically detect clear patterns in the chaos, directing an IT team to a problem’s root cause.
To find the same pattern through even the most rigorous troubleshooting might take weeks, if at all. Teams must first develop hypotheses and then test them, with no guarantee that a root cause will be identified, much less solved. Did we address the core of the issue, or just patch a leaky pipe? Without the right tools, you can never be sure. Algorithms allow the data to take you to the source, often found in unexpected places.
However, there’s still an even larger drawback to traditional problem management: it’s necessarily reactive.
The most seasoned problem manager in the world can only tackle a problem after it has become one; by design, they can’t ever get ahead of the game (save for the notoriously rare Sherlockian hunch). That’s what really separates data processing from troubleshooting — automated analytics need no cues to operate, continuously trolling for even the most innocuous-seeming problems.
Most powerfully, they’re forward looking — software like automated discovery analysis detects problems and notifies management long before services are impacted; predictive analytics suggest steps to avoid future bottlenecks and make services run more efficiently.
Of course, you need smart algorithms to do this. Big data isn’t magic, but the best analytics give that impression because they’re grounded in decades of refinement from top IT mathematics professionals — automated analytic modeling and applied queueing theory aren’t garden variety services. Anything less, and a problem manager may yet be your best bet.
We’re not saying you should forgo hiring the best IT professionals possible — that’s always a priority. But in today’s IT market, they need to be outfitted with the best tools to stay competitive.
(Main image credit: Andrew Hart/flickr)