Managing information technology for a company can be much like painting the Golden Gate Bridge: just when you think you're finished, you have to start painting all over again. This dynamic presents some opportunities and challenges for IT and business leaders.

Among the challenges are these:

• Changes are mostly incremental.
• Benefits/value can be difficult to quantify or articulate.
• Things are done the way they've always been done.

Among the benefits:

• Risks are mitigated by continuity.
• Change takes place in a controlled manner.
• Small improvements can make a big difference.

IT, like painting the Golden Gate Bridge, and like business itself, is a perpetual function that is continuously ongoing and should be continuously improved. Yet some IT functions tend to operate in the here and now, focus just on technology itself, and lack the essential goal of improving IT and business performance.

Perhaps the cause of the here-and-now management paradigm practiced by some IT departments can be explained by the aforementioned benefits. They're too safe. As I read them, I can't help but think "change averse." True, IT should definitely be "risk averse" because of its mission-critical nature. Terms like "uptime," "high availability," "business continuity," and "throughput" come to mind. But IT shouldn't be "change averse."

A good example might be the ongoing debate over COBOL. Many companies still rely on systems with a large COBOL code base. For some, that is just the reality of today and there are practical explanations as to why this situation exists—but it shouldn't exist forever. Why, then, is there a chronic reluctance to sunset aging systems based on COBOL? Reasons often include cost, complex embedded business logic, and lack of skills and resources. But the risks of not renewing systems may be even greater. In today's competitive market, who is willing to bet the future on old technology that is costly to maintain, difficult to change, and functionally inferior? At least there are those safe benefits to fall back on—for now.

Getting back to those benefits, the last one—"small improvements can make a big difference"—actually does hold some potential, especially in change-averse environments. For example, imagine the time savings that were achieved when workers began using sprayers along with brushes to paint the Golden Gate Bridge. That's an incremental change with a big impact. And now bridge engineers are exploring the use of metalized coatings instead of paint, which will extend the time between repainting. They might actually finish the job without having to start over right away!

The same kinds of things can be applied in IT. For example, replacing an older mainframe with a new box can increase capacity, shorten cycle times, open up integration alternatives, and reduce operating costs without introducing risk. That's hardly revolutionary and I'm tempted to yawn, but it's a step in the right direction. Another incremental change would be the use of business process management (BPM) technologies to augment aging systems. While BPM won't replace core systems, it does allow some business logic to be liberated from COBOL, RPG, and so on, and instead be managed in a contemporary, highly configurable environment. And it can get an old shop on the path to modern models, such as service-oriented architecture.

The upshot of all this? Don't let a preoccupation with painting the bridge today keep you from inventing better techniques, better paint, and better equipment for tomorrow. Some of your "inventions" will be incremental, but at least a third of them should be innovative, bordering on radical (think of wireless technologies as an example). And—this is very important—the responsibility for funding and driving innovation falls equally on IT leaders and their business counterparts.

Here are some tactics that have helped others successfully push the IT envelope:

• Focus on sustainable business benefits, not technology.
• Design and build toward the desired state, not just to fight today's fires. For example: Minimize the number of core systems you have and standardize on compatible technologies; reduce the amount of time and resources dedicated to break/fix activities (which requires the root cause to be fixed); invest in non- technology initiatives, such as process improvement and strengthening IT management practices; and consider outsourcing non-core and commodity functions.

Just as the Golden Gate Bridge is in a perpetual state of renewal, IT should always be pushing to improve, reinvent, and deliver more business value. Before you pick up that same old paintbrush, first define your desired state. Get ideas by comparing your operation to industry leaders. Then commit to the management decisions and investments that will make that desired state a reality.