Project Management Keeps IT From Being a Victim of Success

IT departments are either the darlings or the despised of corporate America, and some practitioners would debate which extreme causes the most pain. Let's face it, nowa-days the reward for doing a great job is more work. Once an IT group earns the trust of business units, it must then survive the onslaught of new projects.

Good organizations, like good bosses, don't want you to suck in more than you can handle and implode in the process—our research shows high-performing businesses want IT to stem the tide. And make no mistake: The tide needs to be stemmed. In our recent survey of 986 IT and business managers, 75% said the IT workload was either heavy or crushing.

Poll Results(click image for results gallery)

Smart organizations have gotten a grip on project portfolio management (PPM) and are willing to prioritize and, when needed, put bad projects out of their misery. Like the risk management we discussed last week, PPM is nothing new for mature disciplines. "We've picked up where construction and engineers have been for years," says John Nahm, an IT project manager for the state of Virginia.

In its simplest and most useful form, PPM involves setting parameters around, classi-fying, and prioritizing all projects. When you get right down to it, apart from geeky certifications and expensive software, PPM is something we do every day when we decide to go to work rather than fishing or pick up our children from daycare rather than heading to the gym first. Sure, doing IT PPM right isn't as automatic, but it's not rocket science, either.

Failure By Any Other Name

The highest performers in the IT world, as defined in a recent study by the IT Process Institute, are those most likely to cancel projects—at a rate double that of their lower-performing counterparts, in fact."It's counterintuitive until you think about it," says Kurt Milne, managing director of the Institute. The business world is accustomed to trying new initiatives but being willing to move on if they don't work. But, as Milne points out, in IT there is inherent value in stability, so it's no wonder we're hesitant to pull the trigger. "I don't know that that's a skill that IT folks think they need to have, but logically, it makes sense to shoot your bad projects and move on."

Line of business executives understand capacity planning, prioritization, and making tradeoffs, Milne says, but they expect these choices to be presented in business lan-guage. What they do not understand is when IT overpromises and then under-delivers by failing to meet deadlines or not completing a project properly. Unless IT becomes skilled at PPM, we'll never close this credibility gap.Our survey respondents cited personalities, hidden agendas, and politics as top barriers to efficiently managing their project portfolios. Establishing a good, documented project initiation and ongoing evaluation process is your best defense against managing by whim. Reach out to non-IT departments when you're creating the process so they have buy-in to the final product. When deliberating about the process with those folks, here's what you should be thinking about.

>> Measure work capacity. You can't even begin to justify why you need a process until you have a clear sense of how busy the IT staff is. You must measure. Even if you don't feel your group is overburdened, you can use capacity data in other ways. For example, a project manager in the finance industry told us that he doesn't usually use work capacity metrics as part of the decision as to whether to undertake a project, but they do play a role later.

"We agree to do all requested work," he says. "Then we start negotiating with the customer—or not—about slipping delivery."The fact that 63% of respondents to our poll use "best guess" or don't even attempt to determine their groups' work capacity may also indicate why so many folks struggle with this keystone concept. Of those who do keep track, some use allocated FTE hours, while others standardize on project queue length, that is, the number of simultaneous projects that IT is working on. This works as long as you establish a baseline first, to highlight positive or negative trends as the business environment changes.

>> Build visibility and support. Sharing your data, good or bad, about work capaci-ty is a must given limited staff and financial resources. "It assists IT and business units in prioritizing work and determining if additional resources, like contractors or consul-tants, are required," says a government project manager in California. "It also provides understanding to business about workloads and prioritizing." Or, put another way, how can they feel sorry for you if they don't understand how badly they're beating you up?Reach both up and down the chain of command to establish that your organization needs some way of managing IT project flows. If the C-level people don't get it, you won't get funded for software or project management staff, and you won't be sup-ported in escalating some projects and de-prioritizing, or even killing, others. "We have the mentality that IT does not own anything, meaning that project prioritization is a business processing responsibility," says one CIO at a health care provider. "IT supports the business needs. IT captures, records and reports on requested work efforts. IT management meets weekly with the COO and CFO to prioritize the 'Top 10' projects"

This may be a particularly good strategy for those who would appreciate a "Top 10" rather than a "Top 500" or an "everything is a No. 1 priority" list.

Similarly, if your troops don't get what's in it them, they won't use the PPM tools or processes you fought to implement, and your methodologies will be meaningless. Let's say you establish a process whereby requesters get shunted to project managers if the job seems like it might take more than 40 hours of work. If the help desk isn't in the loop or doesn't buy in, your process fails right at the point of request.

>> Steal globally, evaluate locally. To avoid reinventing the wheel, you should defi-nitely borrow process frameworks from others, though ultimately, project evaluation and portfolio management must be tweaked to fit your culture. If your organization is in a competitive business, try borrowing processes from public entities—after all, your tax dollars paid for them. Several states, including Nevada and Kansas, have public in-formation available about project oversight.

Most assessment processes take into account eight factors. First is fiscal impact, in-cluding ROI: Will it save money or create revenue? Most then add risk, organizational development, customer service, regulatory compliance, health or life safety, board of director strategy or initiative, and loss control. Each may be ranked on a scale from one to 10, then weighted in some manner, with some items (for example, life safety) trumping others, resulting in an overall ranking.When thinking about what evaluation criteria might work for your organization, name your pain. That is, think about what you're trying to accomplish. Are you aiming to decrease closure times on the projects that you do engage in? Twenty-seven percent of the organizations we surveyed do measure this. Or maybe you're trying to better align IT with business needs. Whatever it is, build it into your evaluation criteria.

>> Conquer the wild, wild west. Not everything is a $1.3 million enterprise software project requiring formal work breakdown structures, Gantt charts, critical path analysis, and so on. But neither is everything a quickie work order. "Hey, all we want are a few new scanners, how hard could it be?" can easily turn into a week of delivery, network configuration, troubleshooting, and procurement and configuring of additional licenses for a document management system to accept the scans. There is always middle ground, and for many IT managers, it's this middle ground that is the wild, wild west of IT right now.

"We are spending a lot of time with the small stuff," says Peter Cregger, an application manager in North Carolina. "It's the death of a thousand cuts."

It would be ludicrous to create formal project management overhead for all the small stuff. Yet, without being able to evaluate—and thus control the flow of—all projects, the small stuff will remain torturous. Cregger uses "mini scopes" of work, not to exceed one page, with an abbreviated "roles and responsibilities" document included. This practice captures some of the nutty goodness of PMI's methodology while not using a sledgehammer to perform the work of a tack hammer. The key here is to establish thresholds, be they dollar amounts, estimated work, or both, for what is a small project versus a help desk work order. Use the same types of criteria listed above, though obviously, evaluations that require an investment of executive time are probably not a workable idea. Still, for most of us, something must be done: 33% of poll respondents don't even track small projects, which can't be a good thing.

Just Vote No Trust us, once you get your process in place, the other P-word—politics—will inevit-ably rear its ugly head.

"With budget and resource constraints, someone is not going to get their project done," said an IT manager for a major medical manufacturer. The key question is, will the business unit left out in the cold wait its turn, or try to stage a coup? The magic behind PPM is that, when you do it right, it becomes incredibly clear why a given project shouldn't get done in the context of your overall IT governance strategy. Indeed, some organizations fold project evaluation into their IT governance committees, instead of creating separate teams. Still others, because of the scarcity of executive time, use the team only as an appeal mechanism if the initial evaluation process is unsatisfactory, ra-ther than the primary way that projects get evaluated.

In our discussions with CIOs and IT practitioners, one topic came up again and again: What happens if your portfolio management process comes up with a "not now" or a "no" for a business unit's project, but the business unit—which has its own budget and a degree of autonomy—moves ahead anyway, without IT's approval, and then this rogue project ends up creating urgent unplanned work for IT as the business unit's im-properly planned technology spirals out of control or fails to integrate with enterprise systems?

ITPI's Milne answers this question with a question: "How do you handle it when your corporate strategy says 'We're not going into the Latin America market,' and a line business does it anyway?" If your PPM process is sufficiently integrated into executive corporate strategy, units that are totally out of line will not need to be nailed by IT—the organization will rein them in, with or without IT's participation.

Jonathan Feldman is director of IT Services for the City of Asheville, NC. a contri-buting editor for InformationWeek and Previously he was director of professional services for an infrastructure consulting company providing security and consulting services to the healthcare, financial services, and law enforcement markets. Write to him at [email protected] Levels of PPM Help

All PPM software is not created equal. Some poll respondents rely on collaboration apps, such as traditional project management packages, Microsoft Office SharePoint Server and even wikis, to enhance communications among departments and automate project approvals. These work well for smaller companies and departments, but as you move up the PPM chain, consider more robust, specialized PPM options that will automate the project intake process, integrate with financials, and enable you to do reporting on your entire portfolio's metrics.

Primavera Systems makes the granddaddy of project management software, with suites tailored for specific industries, from aerospace and defense to telecom, but Dap-tiv, a Web based up-and-comer formerly known as eProject, is gaining market share. Other Web-based PPM entries include AtTask and the open-source Project.net. IT groups willing to do some customization should check out RedMine, an open-source cross-platform PPM framework based on Ruby on Rails. Hewlett-Packard's Project and Portfolio Management promises strong integration with asset management. Ec-lipse PPM and CA-Clarity are other offerings with broad integration support.

For smaller shops or those just looking to get started, Microsoft's Groove and Share-Point collaboration systems work well for automating feedback and approvals for projects. MediaWiki is popular open-source collaboration software.

It's important to remember that while good software can make manually updating inefficient spreadsheets a thing of the past, even the best PPM app won't help you if you can't get buy-in for a rational process. In our poll, too many readers cited lack of process and/or political will as reasons PPM failed at their organizations.0