Project rescue

THE DEVIL IS IN THE DETAILS in the discipline of project management—but there is a way out. When it comes to rescuing a project gone bad, vigilant attention to detail can get a project back on track, says David Koenig, chief information officer of the personal market at Liberty Mutual Group. The global provider of insurance products and services is based in Boston, Mass., USA.

Before coming on board with Liberty Mutual nearly two years ago, Mr. Koenig was a management consultant working in a broad range of industries. In this role, he turned around several IT departments. To ensure clients had the proper tools to assess projects, he implemented enterprisewide managerial metrics programs. “Missed or late deliverables, budget overruns and quality problems are common signs of a runaway project. But in most cases, there isn't a single problem or a silver bullet solution to fix projects,” he says.

So Mr. Koenig developed a set of practices based on widely accepted project management methodologies and related metrics to yield a more comprehensive view into large, complex projects. “What gets measured gets done, and you can't fix what you can't measure.”

To this end, Mr. Koenig developed a three-pronged approach to managing projects and rescuing runaway projects:

1 Micromanagement of the project plan

2 Frugal budget management

3 Ruthless management of project scope and quality.

He says these methods can be applied to projects in most any industry, including construction, pharmaceuticals and manufacturing.

4-D Plan Management

All too often, Mr. Koenig asserts, project managers fail to assess how their project is progressing in the broadest context. Steering committees can help with that assessment, especially when they're given metrics in visual form. “A typical project includes a project plan, resource plan and budget,” he says. Visual metrics take that data and overlay a structure using four dimensions to allow for critical components of the project, such as budget, project plan, resource planning and scope definition. By creating graphical charts from these metrics, you're able to look at these areas holistically across multiple sub-projects.”

Mr. Koenig defines four essential dimensions of effective project governance:

Release—Break up larger projects into smaller releases, which are easier to manage and deliver. Another approach includes rolling up projects into a coordinated release. “Set up interim deliverables with tangible measurement benchmarks,” he says.

Functionality Group—Define a logical piece of business functionality from the end user's view. Functionality detail also can be included. “What are they getting? You don't want to say the goal is to deliver a new publishing system. Instead, you want to answer what functionality a group will get when you've completed a release,” Mr. Koenig says.

Team—Identify groups responsible for delivering work elements, allowing managers to separate team-based project plans without losing the end-to-end view across all groups. “Each team can be thought of as a subcontractor. Each has its own project plan and it all needs to come together.”

Phase—Map individual tasks for each team and functionality group through phases of the project methodology used. As an example, use strategy, requirement, analysis and design phases of a software development life cycle. “This dimension allows you to set interim milestones for all the teams toward a delivery release,” Mr. Koenig says.

Frugal Budgeting

Budgeting encompasses both careful allocation of resources and staffing. “Variances in a budget, including resource utilization, are basic and objective measurements of how well a project is being managed,” Mr. Koenig says. “Budget anomalies often foreshadow deeper managerial problems.” He recommends that actual costs versus the budget be discussed and closely tracked at every steering committee meeting or more frequently.

Mr. Koenig cites a couple of common budgeting mistakes. One is simply flatlining the budget and failing to revisit the budget once the project has started. “People say, ‘I have $1 million, and I'll divide it by 12 months.’ Pay attention to how you actually intend to spend the money. Budgets should follow an S-curve, showing resources added in the beginning and rolled off at the end.”

Another mistake occurs when project steering committees fail to hold managers accountable for budget variances. “The biggest mistake governing committees make is when milestones are missed,” Mr. Koenig says. “They often still give the project more money.” Rather, he says, steering committees and business stakeholders do well by acting like venture capitalists. “If you miss your deliverables and can't explain why you have, then don't throw more money at it.”

Taming Scope

Mr. Koenig would like to see quality issues getting the same level of attention as scope, schedule and budget. “Scope management is one of the few areas that steering committees routinely engage in,” he says. “Ruthless scope management requires project managers to force stakeholders to pare their wish lists down to a bare minimum and keep the project team focused on the goals of a particular release.”

PROJECT Rx

Angry customers, bankrupt budgets and late deliverables are merely symptoms of a project gone bad. Instead of focusing on these, examine and correct the bigger-picture mistakes made in initiating and planning that caused the problems in the first place.

David Koenig, Liberty Mutual Group CIO, recommends that project managers examine typical symptoms and probable causes of project shortcomings

SYMPTOMS

• Limited availability of experts and other key resources
• Delays and gaps in meeting business requirements
• IT-led, resulting in IT-focused projects
• Weak partnership between business and IT

CAUSE

Ill-defined business case without clear benefits

SYMPTOMS

• High risk or failure to deliver what business needs
• Poor understanding of project complexity until late in the process
• Last-minute design changes, cutting corners, forcing integration of poorly designed components

CAUSE

“Big bang” delivery approach

SYMPTOMS

• “Scope creep,” the undisciplined changing of requirements as needs evolve
• Resistance to change operational processes as needed
• Lack of focus on long-term viability of the platform or project
• Loss of sponsorship in later phases of the project

CAUSE

Short-term focus on incremental functionality

Quality naturally follows from this less-is-more strategy. “The analogy here is that quality is no different from the tenet of manufacturing,” he says. “You can inspect it in at the end or design it into the product or service. The goal here is to say if you design it properly, all problems downstream are easier to deal with.”

This integrated approach to scope and quality calls for activities to test all elements of functionality and track back defects to business requirements. Particularly in complex projects, such as a major software implementation, it means a willingness to accept a level of non-critical defects while guaranteeing all critical ones are fixed.

Meaningful Metrics

In each of these areas—project planning, financials, scope or quality—metrics allow steering committees to track and oversee projects and escalate issues as necessary. “Projects get into trouble when the project manager or others fall behind and can't catch up, and the project manager doesn't know how to escalate,” Mr. Koenig says. He notes that the project managers may give up, do the task themselves or try to resolve the issue. “Often you can't resolve these problems at a project manager's level alone; metrics give executives a view from the top to lend insight to project managers.”

The steering committee oversees the metrics, while project managers implement them. Essentially, the metrics provide the means to take action and address problems to get a project back on track. “The process doesn't require an investment in new technology or application,” Mr. Koenig says. “It's a different way of looking at common data that's available on almost every project.”

To form the basis for the creation of various metrics, often in chart or graph form, Mr. Koenig recommends any basic spreadsheet application to augment the capabilities of project management software. In addition to sources of project information, including financial, resource management or defect tracking systems, project data are placed into a spreadsheet as a central repository to generate graphical information. “If the data doesn't exist, it's a red flag in and of itself,” Mr. Koenig says. “You can't manage a large-scale project that doesn't have the most fundamental project management information like schedule, budget and staff plan.”

DUE VS. DONE

The due versus done project schedule is often overlooked, yet backlogs can grow out of control quickly and must be monitored closely, according to David Koenig, Liberty Mutual CIO. These graphs offer a side-by-side comparison of what is due according to plan versus what was actually done and when it was actually delivered.

Because of the reporting and administration involved, projects coordinated across multiple groups or divisions of an enterprise work best for creating detailed metrics, Mr. Koenig says. “Projects with a budget of $500,000 and those of five months or greater duration are appropriate for this level of project management.” He says pulling information from various sources can generate a plethora of meaningful metrics. These include a report of late tasks due but not yet completed, burn rate as an indicator of whether a project is progressing as it should, or a schedule of test cases planned versus run.

Mr. Koenig suggests there are three essential metrics for every project:

Due Versus Done—Shows the original project schedule of due tasks against the actual project schedule of done tasks.

Stage Gate Status—Cascades the deliverables and related up-and-downstream dependencies to identify risks to key milestones. As necessary, encourages steering committees to introduce additional resources earlier in the process.

Defects Backlog—A reality check on quality. Shows new and open defects by severity for each team, functionality group and release.

Can This Project be Saved?

Bruce Moore, PMP, national managing principal for project management with RCG Information Technology, has used similar project governance methods and metrics for years on all internal and client projects. “With this unemotional view of reality, it's very quick to see who or what is getting behind,” he says. “You're able to immediately escalate those issues and get them resolved.” RCG is an Edison, N.J., USA-based global IT services firm.

Because project metrics are based on agreed-upon measures, they help everyone understand what's going on in a project, Mr. Moore says. “It's like taking a temperature of the project, and it serves as a great preventative tool to head off problems early on,” he says. “Metrics tell you what you need to know—with the emotional content removed—helping to cut through any laziness or politics that causes people to avoid action. Without these metrics, it's easy to gloss over or dismiss problems.”

Metrics can also reveal the health of a project, as with RCG's implementation of the Livelink knowledge management system. The project represented a considerable capital expenditure for RCG, involving significant resources: An initial project planning team of five in the United States and 13 others worldwide, four business analysts, 10 data converters and 15 others to test the software. “It was important project for us because up to that point we didn't have a single repository for our knowledge management,” Mr. Moore says. “Livelink enabled us to keep all the data and project artifacts in one place and give access to our consultants to that accumulated knowledge.”

With Livelink, metrics around two key areas—tasks planned versus actual and hours planned versus actual—told the good news. RCG IT was able to deliver the project in three months, instead of a planned six-month period. “We didn't reduce any staff resources on the project though,” Mr. Moore says. “All our staff could access all collateral on all projects and methodologies earlier on so it was a great benefit to finish early.”

Even with the best processes and metrics in place, a lot boils down to the individual project manager, according to Mr. Koenig. “There is no substitute for good management, and no methodology that can overcome bad management,” he says. “Before anything else happens, good project managers need to be in place.” PM

Marcia Jedd is a Minneapolis, Minn., USA-based supply chain and business writer.