In defense of IT dashboards
It is impossible to manage effectively without the metrics to benchmark and shape performance
- By Robert L. Otto
- May 06, 2010
Robert L. Otto is an executive vice president at Agilex Technologies. He worked in the federal government for nearly 40 years, culminating in the position of chief information officer/chief technology officer at the U.S. Postal Service before his retirement in 2007.
The Obama administration has made a significant commitment to the use of performance metrics as a strategic management tool. However, some people have derided its focus on information technology dashboards as a short-sighted, simplistic response to the complex problem of managing government agencies. As a result, they are prepared to discount the insight that those leading indicators can provide.
It's similar to the situation I faced on being named chief information officer at the U.S. Postal Service in 2001. At that point, we had little data regarding how we were performing as an IT organization, and management decisions were often based on gut-level intuition. However, by moving to a metrics-driven approach, we were able to quickly transform IT performance while cutting IT spending in half and dramatically improving client satisfaction.
Early on, I recognized that it's impossible to manage effectively without the metrics to benchmark and shape performance. In our case, we rarely made explicit promises because we had no way of tracking our execution. As a result, the perception of IT was lackluster, at best.
To change both perception and reality, I created a war room/operations center in which dashboards that detailed our day-to-day performance were prominently displayed. By doing so, we sent a clear message to everyone we met with — internal clients and external partners — that we intended to be held accountable, and we were using rigorous, business-level benchmarks to manage ourselves. Over time, the approach helped people believe that we could be a valued partner in improving the organization’s performance.
For example, help-desk performance is something that affects everyone. By making it a key metric for my team, we were able to systematically reduce the call volume from 150,000 to 25,000 a month. We also measured and improved responsiveness, completion rates and user satisfaction to ensure that we were not overly biased by one metric.
By implementing a metrics-driven approach, we realized two other significant benefits. First, it prompted us to closely examine our business architecture because we needed to determine how one set of metrics influenced another. As with a balanced score card, it helped us understand how fairly mundane things could have a cascading impact on enterprise performance. In many cases, that insight subsequently underscored the business case for making necessary infrastructure improvements.
Second, our desire to make meaningful, like-for-like comparisons drove us to be relentless in our efforts to standardize, consolidate and simplify our operations. If we couldn’t measure it, we probably couldn’t manage it. As a result, we needed to redesign, replace or eliminate the outliers. By streamlining our operations, we were able to reduce our operating costs significantly while dramatically improving the reliability of our enterprise systems.
I should add that the dashboards we started with — which were manually printed reports — were fairly rudimentary compared to what’s available today. One of the benefits of using the latest data visualization technology is the ability to work collaboratively with various stakeholders to weigh the pros and cons of specific scenarios.
As we have seen, many people will point out the shortcomings of IT dashboards while ignoring the need to use performance metrics in decision making. If our reporting tools fail to provide the insight we need, we need to improve and expand the dashboards accordingly. But whatever you do, manage objectively, which means managing by the numbers.