Break out of the boardroom
NATIONAL GRANGE MUTUAL INSURANCE CO. PULLED TOGETHER AN EXPERIENCED CROSS-FUNCTIONAL TEAM TO BRAINSTORM ITS 2004 STRATEGIC PROJECT PLAN.
BY MARK LALOMIA, PMP, AND PAUL FREITAS, PMP
NATIONAL GRANGE MUTUAL INSURANCE CO., Keene, N.H., USA, used to make last-minute pushes to complete the coming years' business planning. Members of the information technology (IT) management team independently brought their clients' projects, pet projects and personal agendas to the planning table. Discussions included a blend of horse trading, jockeying and stiff negotiating.
Over several days, which sometimes turned into weeks, the team came to some consensus on which projects should get completed and who would work on them.
This year, the team found a better way. The program control office (PCO) would determine how many of the firm's client projects could get started and, more importantly, completed by the end of 2004.
National Grange Mutual Insurance Co. needed a more efficient way to prioritize and schedule its information technology project queue.
The company's program control office established a task force to review each project's scope and necessary resources.
Although director buy-in was difficult to obtain initially, the task force was authorized to proceed with a pilot project planning session.
Based on initial success, the team was given five days to plan as many projects as possible, and the result was a well-received, high-level schedule and resource allocation plan.
The PCO staff “white-boarded” objectives, strategy and a rough plan it believed would produce reliable and accurate information through a sustainable, repeatable process.
The team gathered 50 client-requested projects of varying size, scope and complexity using a portfolio management process involving the IT steering committee. These projects, each characterized with a scope statement, would be prioritized into one top-down list.
Because National Grange Mutual Insurance's IT department also maintains existing systems and infrastructure, the team had to decide how much of the IT department would be allocated to maintenance work. To accurately determine the department's capacity, IT management would assess the skills of each team member, and these skills would be documented in an easily accessible spreadsheet.
A task force would represent all areas needed on the projects to be planned. It would be seasoned, outspoken and able to make decisions with limited data. It needed to be a cross-functional team from the IT department that included two business analysts, two developers, a systems engineer and a quality assurance person—each with different domain knowledge representing most of the projects to be planned.
A TASK FORCE WOULD REPRESENT ALL AREAS NEEDED ON THE PROJECTS TO BE PLANNED. IT WOULD BE SEASONED, OUTSPOKEN AND ABLE TO MAKE DECISIONS WITH LIMITED DATA.
Last, each of these components—the task force, prioritized project list, skills and department resource list—would be used to develop high-level schedules for the 50 projects. The PCO would use the schedules to determine how many of the projects could be completed based on the capacity of the department.
The first challenge was to get the chief information officer (CIO) and IT director team to buy into the proposal—not an easy task because the proposal shifted control of planning projects and assigning resources. As expected, the IT directors were not thrilled about the loss of direct control, but the plan was given a chance.
The director group agreed to a one-day pilot in which it would be able to review the output. The PCO selected three projects that fairly represented the department's work. The task force reviewed the project statements and came prepared to discuss the projects.
The task force pilot meeting began with a discussion of what each project was expected to accomplish. In most cases, the project statement provided enough information for the task force to understand the project's scope, but the team was prepared to pull any department member into this meeting if necessary to explain in greater detail.
As the project scope discussion went on, the team made assumptions that were captured and documented for future reference. A predefined set of tasks using a Microsoft Project file was used as the input template for the schedule. This template, which fed all of the high-level project schedules, was displayed on the wall for all to view.
Next, the task force determined the “right” tasks that would allow the department to meet the project goals. Using the template, the team added or removed tasks as necessary for a particular project and determined task interrelationships.
To ensure consistency, the team used the program evaluation and review technique (PERT) estimating model. Every member of the task force produced a pessimistic, optimistic and most-likely duration estimate for every project task. The project template was reviewed task by task with members sharing their estimates in a round-robin fashion. These raw estimates were collected in a spreadsheet, which averaged the raw numbers and plugged them into the PERT estimating equation. This model could take three estimates from each person (pessimistic, optimistic and most likely) for each task (18 estimates total) and determine one estimate for the task.
The process was clunky at first, but the team quickly caught on. The project schedule then was reviewed task by task, and the necessary resources were assigned to the tasks. If the team felt a specific person was needed for a task, that person's name was placed on the task. If any person with the skills could perform the task, a general role was placed in the task, such as Java developer or testing analyst.
It took about an hour to understand the project scope and two hours to build the high-level schedule—three hours per project.
The next day, an interesting thing happened: The task force met to determine how to handle the post-pilot meeting with the IT directors. In this meeting, the biggest advocates became the task force, while the skeptics became the biggest allies with the loudest voices. It was decided to let the task force speak instead of the PCO.
While the IT directors still were apprehensive about relinquishing direct control of planning and assigning resources, they could not ignore the task force. With the support of the CIO, the task force was granted five full days to plan as many projects as it could.
In their review, the directors suggested that the task force also determine a confidence factor for each project. When they completed project planning, they voted individually on how confident they felt about the project estimate. Predefined scoring sheets were used to determine an overall confidence factor for the project; every task force member filled one out for each project.
A five-day planning session was scheduled, the projects were queued up, and the task force went to work. The process went smoothly—the task force planned 27 projects and three additional projects that surfaced during their efforts. The biggest challenge was getting the documented project statements from the appropriate people. If a project statement was not received, the IT director or project manager attended the meeting to discuss and “white-board” the scope of the project.
THE NATIONAL GRANGE MUTUAL INSURANCE TASK FORCE
Top Row, From Left Rob Miner, Karen Maynard, Kelli Bleau Bottom Row, From Left Jeannine Lawson, Mark LaLomia, George Roy, Rose-Ann Ciercielli Not Pictured Paul Freitas
There were 10 projects spanning 2003/2004 that did not need to be reviewed. The existing detailed project schedules were used to determine resource allocation of the spanning projects. Allocations existed for 40 total projects (of the original 50), and the PCO used this information as input to the resource model. While building the resource model, the team took vacation, sick and education time into account.
Through the process, the PCO determined that the IT department had the capacity to do all of the projects the task force planned. Additional intranet and infrastructure projects could be taken on. The team didn't want to over-commit the department's resources while making the most efficient use of them. Commitments made to clients needed to be honored, while keeping the department busy with challenging work.
Following the five-day session and the development of the resource model, the PCO staff communicated the findings to the CIO and the IT director group. This meeting went smoother than before, as everyone was getting more familiar with the new planning process.
THE TEAM DIDN'T WANT TO OVERCOMMIT THE DEPARTMENT'S RESOURCES WHILE MAKING THE MOST EFFICIENT USE OF THEM. COMMITMENTS MADE TO CLIENTS NEEDED TO BE HONORED, WHILE KEEPING THE DEPARTMENT BUSY WITH CHALLENGING WORK.
As a result of the successful effort, the task force was asked to plan five additional projects not addressed within the five-day session. The PCO scheduled follow-up meetings with individual directors to review their accountable projects, the project assumptions, high-level schedules and the resource plans for people within their direct chain of command. The PCO staff was looking for director feedback and buy-in on project start times, resource assignments and priorities.
The new planning process utilized existing expertise and created the synergy National Grange Mutual Insurance needed to streamline the difficult business of scheduling projects into the future. Moving forward, the PCO expects to be provided with additional projects or shifts in project priorities. The process wasn't easy, but in hindsight, it was the right thing to do.
Over the next year, the PCO hopes to continue its high-level planning and resource allocation and will measure and monitor the actual results against the task force's assessment to refine the process. PM
Mark LaLomia, PMP, and Paul Freitas, PMP, both are program control office managers for National Grange Mutual Insurance Co., Keene, N.H., USA.
AUGUST 2004 | PM NETWORK
PM NETWORK | AUGUST 2004 | WWW.PMI.ORG