Turning metrics into money

Compute power is like horsepower: just as auto enthusiasts are eager to squeeze a few extra ponies from under the hood, so too do IT junkies relish the thrill of an upgrade, whether it's a faster server or a new software architecture that promises to let them reclaim their weekends.

Sadly, those in the bean-counting centers don't necessarily share this love of IT power (they tend not to love fast cars, either -- witness the number of the Toyota Camrys in the parking lot). In fact, given the excesses of the '90s, when "info-tech binging" became de rigueur, the state of CFO-CTO relations has never been more hostile.

All of this makes it increasingly difficult to push through much-needed IT budget expenditures. Fortunately, modern IT professionals have a secret weapon: APM (Application Performance Management) tools that allow them to accurately quantify the performance characteristics of their run-time environments. By leveraging the depth of information provided by these tools and applying a few common sense analysis techniques, you too can win over the enterprise actuaries.

The first step is to carefully define your metrics. For example, a major financial services company recently faced a budget crisis: Their trading-floor staff needed high-powered workstation hardware, but they were getting pressure from accounting to save money by investing in high-end value PCs instead.

Using APM tools to instrument their current trading floor run time and translate the recorded CPU utilization data into a more tangible metric, the IT staff figured out the delay, in seconds, incurred during peak usage. For a company where time is literally money, and where even a few seconds' delay could mean the difference between profit and loss, the hard data elicited a more favorable response from accounting: workstations all around.

Of course, not every situation is as cut-and-dried. In some cases, you will need to aggregate several metrics to paint a compelling picture for the CFO. The old stalwart, TCO (total cost of ownership), has been flogged to death; the bean counters already view IT as a cost center. Proposing ways to reduce your department's drag on company profits does little to fire their imaginations.

Instead, your best bet is to focus on how a given outlay will positively impact the bottom line, using APM data to back up your message. If you can demonstrate how your proposed upgrade will allow the company to do more without increasing head count, you can write your own check.

This is where many of the internal APM techniques come into play: orchestrating critical business processes and then monitoring their performance across both baseline and upgraded scenarios. When your APM data shows that xnumber of workers can deliver ymore widgets per measurement period (and when the outlay figure is dwarfed by the potential profit increase), victory is close at hand.

In the end, leveraging APM is all about common sense as it relates to profits, productivity, and the bottom line. Much damage was done during the raucous '90s, both to corporate profits and IT's reputation as a responsible business function. Your mission is to change the perception of IT. By shifting the discussion from cost-centric to profit-centric themes, you'll find your proposals will be met with less resistance and that the rift between the number crunchers and the IT team will start to heal.