Earned value management
- 06 April, 2006 13:11
If your IT shop isn't using earned value management, you may want to start thinking about it. EVM, which has its roots in the U.S. Department of Defense, is moving into private industry. More important, it's coming to IT.
When used properly, EVM helps team members, project managers and their bosses accurately gauge progress against an established project plan. EVM also enables teams to accurately assess where they'll be in the future, allowing managers to make key decisions on resource allocation or revisions before projects unexpectedly spin out of control.
"When people report percent complete, you might get 'I'm 50 percent complete,' based on that person's intuitive knowledge. Earned value analysis takes that guesswork out of it," says Robert Leto, director of the IT effectiveness practice at PricewaterhouseCoopers Advisory Services in New York.
The Defense Department has employed EVM for years, requiring its contractors to use it for reporting on federal projects.
So, what exactly is EVM? "It comes with a reputation for being complicated and difficult, but I don't think anything could be farther from the truth," says John M. Nevison, president of Oak Associates a Mass.-based firm that provides consulting and training services related to project management.
EVM is based on several figures that are used in calculations to determine whether a project is adhering to schedule and budget. Results can be measured in terms of money or time.
EVM is not about producing perfect scores. "It's accepted that you're going to vary from your plan," says Marilyn S. McCauley, owner of McManagement Group, an EVM consulting and training firm in Ohio. "If I see [perfect scores] every time, someone's cooking the books, because that's not reality."
The point of EVM, she says, is "to see how close we are against what we planned, and when we're not close, to ask, 'Why aren't we there, and what are we doing about it?' "
This is where EVM offers much of its value. If project managers and their executives can see early on that projects are falling behind schedule or going over budget, they can make key decisions about how to proceed, rather than reacting to problems after the fact.
"What earned value does is provide you navigational tools early to let you look ahead to see where you'll be if you do nothing," Nevison says.
Earned value calculations can be done at various points during a project, but the numbers tend to stabilize when you're about 20 percent through, says Quentin W. Fleming, co-author of Earned Value Project Management (Project Management Institute, 2006) and a management instructor at the University of California, Irvine.
"So the point is, if you're 20 percent through the project, you can predict what the final costs are going to be, plus or minus 10 percent," Fleming says. "It's a very powerful tool, and here's what's powerful: If you're 20 percent through a project and you've been authorized $1 million, and your cost efficiency to date suggests you're going to need $2 million to finish the project, then management has decisions [to make]."
Despite EVM's reputation for offering insight into project progress, many IT executives aren't yet embracing the discipline, partly because the underpinnings that make it work aren't in place.
"The concept of earned value is really elementary project management. But the problem with many IT organizations is that they don't use rigorous project management methodology," says Dan Gingras, a partner in the information technology leadership practice at Tatum, a consulting firm in Atlanta.
"In order to do earned value management, you have to have a good budget," he says. "In order to have a good budget, you have to back up one step further and follow a good project management methodology."
Project managers and IT leaders need to accurately define a project's scope and requirements, develop specific work packages and work breakdown structures, and then establish a good budget, says Gingras, who also teaches technology strategy and system design as an adjunct faculty member at Boston University's Metropolitan College.
Moreover, Gingras and others say that employing EVM in projects takes plenty of training. There are books and multiday courses that teach the practice.
So, why go through all this upfront work? Gingras points to a well-known fact: "Significant numbers of projects aren't completed on time or on budget, or they don't deliver what they're supposed to deliver."
EVM can help improve your chances of project success, says McCauley. "The idea behind earned value is seeing what you need and when you need it. It's all about management, and it's all about control."
Who's driving the EVM bandwagon?
Earned value management is beginning to catch on, says Quentin W. Fleming, a management instructor at the University of California, Irvine, but many in IT resist it because it would require changes in how they set up and manage projects. For example, IT shops would have to thoroughly define, scope and budget their projects in advance rather than employ incremental development, as many currently do, Fleming says.
Despite this reluctance, CIOs may find CEOs and CFOs asking them to use EVM as a way to bring more transparency to projects, says Robert Leto, director of the IT effectiveness practice at PricewaterhouseCoopers Advisory Services. "Will the IT community embrace it? I don't know if they will, or if they'll be mandated by the boards or the CFOs," he says.
Pressure to use EVM in IT is also coming from the federal government. Marilyn S. McCauley, owner of EVM consulting and training firm McManagement Group, says the Sarbanes-Oxley Act's requirements for greater financial transparency are becoming a driver for EVM. And the president's fiscal 2007 budget calls for IT departments at more government agencies and organizations to apply EVM.
"It's becoming a requirement in IT industries that deal with the government," says McCauley. And that, she says, will have a ripple effect that could spread well beyond federal departments and contractors.
Using EVM in IT, McCauley says, "is not a matter of if; it's a matter of when."
An EVM primer
Earned value management is based on several figures that are used to calculate a project's progress. You can measure in dollars or time.
Planned value (PV): This is the value of all resources needed to do the work to meet the project's objective. Although most project managers calculate PV in dollar terms, some calculate it in terms of time -- the number of hours it's expected to take to complete the project.
Let's take a very basic example. We've budgeted $200 to buy, set up, network and test a new system. We've budgeted $50, $75, $50 and $25, respectively, in materials, labor and other costs for those four phases.
Keep in mind, though, that the $50 set aside to buy the system doesn't just cover the cost of the actual hardware and software. It also takes into account the value of time that will be required to find the right system, the time that will be needed to fill out the purchase orders, the time it will take to actually buy the system and so on.
"The basis for earned value management is worked performed, not money spent," says Marilyn S. McCauley, owner of McManagement Group, an EVM consulting and training firm. Our PVs are $50, $75, $50 and $25.
Budgeted (cost) at completion (BAC): This is the sum of all PVs -- the total for all phases. In our example, BAC is $200.
Earned value (EV): As our team completes portions of the planned work, we check off that work and the amount of money (or time) it should have taken to do it according to the project plan. Project managers calculate EV at predetermined times based on the plan, typically at the end of the company's accounting period, McCauley says. We've completed Phase 1 -- buying the system -- within the planned time frame. Check that off as done. Our EV is $50.
Actual cost (AC): This can also be measured in dollars or time. In a perfectly executed project, EV and AC are the same. But in our example, let's say we actually used $60 in resources to buy that system. Our AC is $60.
Once you have these figures -- PV, BAC, EV and AC -- you can calculate other numbers that tell you about your progress on a project. Here are some of those calculations:
Schedule performance index (SPI): EV divided by PV for a particular phase of a project. In our example, that's 50/50 = 1, a perfect score for Phase 1, indicating that we're on target for schedule. "I said I'd do $50 worth of work, and I did $50 worth of work," McCauley says.
Cost performance index (CPI): EV divided by AC. For our project that's 50/60 = 0.83, indicating that we're underperforming for our costs. "For every dollar I'm spending, I'm only getting 83 cents worth of work," McCauley explains.
In a perfect project, the answer is 1. But most projects fall below that because most projects miss their targets.
Estimated (costs) at completion (EAC): BAC divided by CPI. The answer is a forecast value in either dollars or hours that indicates the projected final project costs or time. There are various formulas for EAC, McCauley says, but this is one of the easiest to use. In our example, that's 200/0.83 = 240.96. This indicates that at the rate we're going, the final cost will be $240.96 rather than our planned $200.
Schedule variance (SV): Subtract PV from EV. In our example, our earned value is $50 because we've done the first of our four phases: We bought the system. The PV for that first phase was actually $50. So 50 - 50 = 0. That's a perfect score, so we're on schedule.
Cost variance (CV): Subtract AC from EV. In our example, that's 50 - 60 = -10, indicating that we've overspent by $10. If we were on target, CV would be zero.