The art of gauging risk

They're often called war rooms or, perhaps more ominously, murder boards — they're the places where government information technology managers hash out real and potential risks facing their projects so that they can reduce the chance of failure.

An increasing number of program managers are trying such innovative techniques to better identify and alleviate project risks. However, despite the growing popularity of these efforts — some hatched by internal staffs and others based on available risk management frameworks or tools — many program managers still fall prey to risk management pitfalls, industry experts say.

The most common mistake may be failing to formalize risk management efforts by not employing and documenting specific strategies. Other missteps arise when IT project leaders fail to convince the entire team that risk management is everyone's job, now a popular concept called enterprise risk management.

And believe it or not, some people's aversion to math is making matters worse. Many project managers balk at the painstaking practice of quantifying risk factors and mapping them in spreadsheets.

"We must have informed project managers. But the level of innumeracy in all of us and in organizations is quite high, and many people are hostile to quantitative approaches," said J. Davidson Frame, academic dean of the University of Management and Technology (UMT) and author of the 2003 book "Managing Risk in Organizations: A Guide for Managers."

Despite the shortage of risk management skills in government, most agree that new national security concerns and growing budget deficits have underscored the need for risk management as a way to increase the success of IT programs.

"Risk management is such a popular topic now that everybody says they are doing it," said Adrian Bowles, an analyst at the Robert Frances Group Inc. "The conclusion they've reached on vulnerability is that it is like regulation: It's not going away."

Project risks, however, can easily outlast a project manager's attention to manage those risks. "Risk management can materialize after an emergency and fade out," said Tony Maturo, head of NASA's Academy of Program Project Leadership. "However, risk is an everyday occurrence that we as an organization must use, think about and analyze."

Know your foe

War rooms and group brainstorming sessions designed to ferret out potential program risks amount to risk identification, the first step in any risk management exercise.

Because risk identification is rooted in common sense, even less formal risk management efforts — the seat-of-the-pants approach — begin this way. "Everyone sits in a room and asks, 'What is going to bite me?' When the group figures out what that is, you write it down and stick it in a drawer," said Bob Overstreet, director of GRT Corp.'s Program Management Office.

But a more structured and rigorous risk management effort can keep those problems from being banished to desk drawers, added Overstreet, who spent years as a military program officer. "You go back with a structured process and plan of attack," he said. "This starts by ranking risks based on probability and potential impact to the project."

Risk rankings help managers decide between the options they have when encountering risk. According to textbook program management principles, officials can decide between risk transfer, risk mitigation, risk acceptance and risk monitoring. For instance, a risk can be transferred through the use of a manufacturer's warranty. Or it might be mitigated by finding the cause of a problem posing the risk and fixing it, UMT's Frame said.

Then there is risk acceptance. "This means more than just lying in a fetal position sucking your thumb," Frame said. Instead, risk acceptance is the process of walking through the possible consequences of an identified risk and actively deciding that mitigation costs outweigh the risk.

An exaggerated example of risk acceptance was a decision city officials from Tampa, Fla., made after using Digital Sandbox Inc.'s Site Profiler, a risk analytic application (see box, Page 26). Officials used the software to run through all of the possibilities for safeguarding City Hall in the wake of the Sept. 11, 2001, terrorist attacks.

"It's obvious that officials were not going to build a fortress around City Hall to stop any damage that could be done with a truck outfitted with explosives," said Anthony Beverina, the company's president and chief operating officer.

Although the conclusion seemed obvious, Tampa's decision was grounded in data. "Our planning and management department was responsible for meeting our mayor's requirement that our risk analysis be empirically based," said Rick Smith, Tampa's planning and management director.

Finally, risk monitoring involves actively following a targeted risk. "The idea is to reduce an area of risk from high to moderate and continue to mitigate it until it is a low-impact item," Overstreet said. "Once you get the risk down to low impact, you don't keep working on it. However, the project manager should keep it on his list and pull that out every month or two to make sure it is still low priority."

The level or nature of program risks naturally dictates which of the different methods of risk response a program official will employ. "In government, there are four levels of risk [schedule, finance, return on investment and human factors]," Maturo said. "But especially for NASA — which contends with the risk inherent in exploration — human life is top priority."

Aside from safety-related risks, other major risk categories include the trio of bugaboos that plague government and private industry: schedule, finance and return on investment. "Risk, simply put, involves looking at the probability of an event and the cost of that event if it comes to fruition," said Tom O'Rourke, a senior consultant at Total Quality Organization.

Enterprise risk management

Most risk management experts agree that every individual on a program staff should share in risk management efforts. "Risk is the responsibility of everyone. By saying that, we mean that it starts with the lowest level, the individual contributor, who must go through the same kind of steps as those on the program office level," Maturo said.

Some program management experts liken this new theory of enterprise risk management to the 1980s trend of total quality management, in which everyone is involved in improving the enterprise and, therefore, services and products.

But that idea has met with some skepticism. "I don't see it happening," Frame said. "I don't think it is practical. It sounds good on paper, but when you try to implement it, you run into trouble."

Risk experts say spreading the responsibility of risk throughout an organization, instead of assigning ownership for each potential risk, is a problem. "You assign not a group but an individual — a champion — to work on a particular risk," Overstreet said.

Officials at consulting firm Robbins-Gioia LLC also instruct agency clients to be specific on risk responsibility. However, the firm warns against letting top officials off the hook.

"I think we are seeing more people with the title risk manager and usually they are set up in the program office," said Keith Kerr, a senior consulting manager at the firm. "However, you have to be careful how you assign risk. Both the program manager and the functional manager in an organization have to own risk."

Heightened awareness around risk and responsibility, however, does not have to stifle new ideas and projects. And rigorous risk assessment doesn't necessarily lead to finger pointing and a hostile environment, experts said. "Risk management can impact the politics of an organization in a positive way, since you are increasing communication and providing insight at all levels where the impact of risk might occur," NASA's Maturo said.

Jones is a freelance writer based in Vienna, Va.

***

Risk management warning signs

If your project has any of the following attributes, it may be at risk of failure. Watch out if your project team:

* Has not identified a specific list of risks it will actively work on.

* Is not fully aware of the more common program risks.

* Has no framework in place for systematically identifying and addressing risks.

* Does not understand the project's endpoint or deliverables.

* Has limited interaction with operational or business staffs.

Source: GRT Corp.