Evidence-based management is an approach where hunches are discarded and instead decisions are based on hard facts. It’s pounded into aircraft pilot training: when there is no visibility, regardless of what senses are suggesting, pilots must trust their flight instruments. In medicine (where the practice of evidence-based decision making originates) new diagnostic technology can trump the wisdom of physicians and can make decades of experience only an input into a course of treatment and not the final call. With few exceptions, good data is the best way to make great decisions. Without it you’re essentially flying blind.
We use evidence-based management all the time in business. Historically, we’ve just called it something else. In business we’re consumed with data and metrics. We make decisions with data and we measure performance with metrics. It’s letting the data and metrics — the evidence — tell the story and then taking some form of action on it.
As a business function, we can all agree that internal IT is replete with data and opportunities for metrics. Even still, it is surprising how poorly IT organizations measure what they do and how they make decisions. In my firsthand experience the problem can be attributed to at least three characteristics:
- Lack of recognition for the value of data and metrics.
- Insufficient skills in determining how and what to measure.
- System limitations/issues.
The effective CIO must address these and knock all three out of the ballpark. The success of an IT strategy is predicated on good metrics. Clearly that’s easier said than done. What every IT leader quickly learns is that without even a basic set of metrics, management of processes is close to impossible (at least in any quality manner) and making an argument for a course of action, particularly to the boss, is exponentially more difficult. To paraphrase Druker, who I believe got it exactly right, if you want to manage something you had better measure it. That’s an important rule.
Choose wisely if your metrics are currently far and few between. Go after the most valuable items, but keep the list relatively short. Produce a long list and you risk pushing your team into a tizzy. If you inherit an environment that appears to monitor and measure in excess, find a way to reduce it to a list where each item has merit. Simply asking the purpose of each metric will truncate the list quickly!
I’ll be the first to admit that metrics aren’t the most glamorous part of IT leadership, but they must be a priority. If leveraged in the right way, quality metrics can be the difference between dysfunctional operations and high-performance. That’s a golden rule that any CIO should want to follow.