Home > Articles > Programming

  • Print
  • + Share This

Developing a game plan

Your process group has been devastated by turnover since you reached Level 3 two years ago. Last year, management cut the group's budget in half when it was trying to reduce overhead expenses. Then three of the five remaining engineers left to work on projects because they felt that process group assignments were parking positions. The best people in your firm seem to be put to work on projects, not overhead assignments. That's where the promotions are and what's considered fast-track. Besides you, the only people who are left in the group are two retirees who were hired as consultants to work with projects because they have credibility and two part-time courseware developers. Fortunately, you have just received permission to hire an analyst to help implement your Level 4 metrics requirements. To get the group back on track, it must be reinvented and revitalized (its mission and people's perceptions of its role must be changed). In addition, the group must be staffed with high achievers who can generate results quickly.

Besides you, the process group has four people filling the eight slots that are illustrated in the organization chart in Figure 5.4. As mentioned, two retirees are on board and acting as liaisons with projects. These two senior practitioners are held in high regard by the working troops. Your other two employees are part-timers from academia whom you brought on to develop education and training materials. The candidates for the two process developer vacancies are well-thought-of people who are between assignments. The third vacancy will be filled with the new metrics hire. Your task is to develop a game plan for reaching Level 4 as the team is being built and middle management is being convinced to support it.

Figure 5.4: Process Group Organization Chart

The budget for the process group for this year is $2.4 million:

  • Personnel (four employees) $700K
  • Consultants (two retirees) $450K
  • Academicians (two part-timers) $200K
  • Assessment support $200K
  • Education and training $250K
  • Promotion and outreach $250K
  • Specialized Web tools $100K
  • Web site development $250K

The budget funds the staff's developing and fanning the process out to project organizations ($1,350K), training ($250K for seminars brought in from the outside), promotion ($250K to prepare a newsletter, work with the customers, and attend conferences), and assessment support ($200K to bring in an external assessment team). The major new task to be pursued by the process group is Web site development ($350K). The group will put a process asset library on line to make their products (processes, training materials, improvement metrics) easily accessible to those who have access via the firm's network. They will continue tracking their performance using the metrics data on defects and costs that they collect as part of their process.

There is lots of pressure on you and the process group to move to Level 4 within two years. In its zeal to support the process improvement initiative, senior management took your suggestion to make process one of the factors in its middle management salary bonus scheme. Because their paychecks will be directly affected by the move to Level 4, many middle managers have suddenly awoken to the importance of process. They have started asking questions about what's involved in software process improvement. Unfortunately, the burden of success or failure falls on you because management cooperation with the process group isn't listed as one of the criteria for use in computing the bonus.

You are assuming the role of the process group manager for this exercise. Assume that you have just been promoted to fill the position. How do you get started? Do you staff the group or plan first? Who should get involved; what should they do? When do you kick off the effort? How do you build and energize the team? Where do you find qualified people for the group? What are upper management's expectations? Which program managers support you? Will they collaborate with you and permit their programs to serve as pilots? Senior management has summarized these questions in one: "What are your plan of action and milestones?" Middle managers ask a different question. They want to know, "What's in this initiative for me?" In other words, they are more interested in the business case justifications than in plans for moving to Level 4.

Before developing a game plan, some factfinding should be conducted. What is the group really being asked to accomplish? Is Level 4 the game, or is the game changing the culture as discussed in Chapter 1? Do you know? Have quantitative objectives for the effort been finalized and agreed on? If not, seek clarifi-cation. What are the measures of success for the effort, and how will they be demonstrated? How do these measures relate to the goals set for the effort? Will an outside observer be required to confirm that you've reached Level 4, or can you use someone from another division? Again, if you don't know, you should find out. You need to dig deep to discover hidden agendas.

Luckily, the real objectives have been defined using the goal–question–metric (GQM) paradigm discussed in Chapter 2. As shown in Figure 5.5, these goals and their related measures of success focus on productivity improvement to justify the expenditure of $2.4 million annually to pursue process improve-ment."Where did this number come from?" you are probably wondering. When you dig, you find out that the numbers were developed by projecting productivity gains achieved three years ago into the future. Because the numbers are being questioned, you will have to determine their merit as you develop your plan and business case.

Figure 5.5: GQM Worksheet

You can start your plan effort in earnest. Because you're relatively new to process improvement, you have brought a management consultant on board to guide you through the planning process (for at most six months). Your boss has recommended that you pay for this expert with surplus funds. Your strategy here was to avoid political infighting by providing someone credible whom everyone could agree with (or blame) to help craft the business plan/case. The process the consultant recommends following is shown in Figure 5.6.

The process in Figure 5.6 homes in on steps 4 through 7 of the seven-step business planning process discussed in Chapter 2 (develop a business plan, prepare the case, sell the idea, and execute) (see Figure 2.4). The process in Figure 5.6 starts with vision and translates it into the work that needs to be performed via goals using activities to group tasks that are similar. Let's briefly discuss what needs to be done during each of the steps.

Start the Process by Involving Stakeholders

Start by establishing an infrastructure aimed at getting stakeholders involved in the planning process. Charter a steering committee and working groups. Invite key people and influence makers to participate. Hold your kickoff meetings as soon as you can. Build group consensus. Publish and promote the results via your Web site. Address the legitimate concerns raised by these groups. Show participants that their inputs are important by acting quickly on their suggestions. Identify stakeholder "win" conditions, and plan to use them to prioritize your work accordingly [Boehm,1998].

Figure 5.6: Recommended Planning Process

Develop a Top-Level Vision and Strategy

Building on the discussion in Chapter 2, write down what the initiative is trying to achieve (vision) and how you are going to make it happen (strategy) in simple language that anyone can understand. You will be surprised how hard it is to write such a document. Once you have your ideas on paper, review the document with your stakeholders, improve it, and have them take ownership of it. This document, which should go on your Web site, will help you explain the initiative, its goals, and how the goals are going to be achieved to anyone who asks.

Define the Work to Be Performed

Based on your GQM objectives, define the tasks that need be done in the near and long term. The work breakdown structure shown in Figure 5.7 organizes the work that needs to be performed to achieve these goals using activities and their subsidiary tasks ( Figure 5.7 shows the mapping between tasks and activities). For each task, define the inputs, outputs, deliverables, and dependencies. Don't get bogged down in details when you start. Begin by identifying the most important things you must do to be successful. Cut the list to the five most critical things you need to do using stakeholder inputs and win conditions to establish priorities. For example, statistical process controls must be deployed to achieve Level 4 (task 3.3). But you are not yet staffed to handle this challenge. You could correct this problem as you develop your plans by filling one of your vacancies with someone with the skills to tackle this task. The person you hire in turn can act as a mentor and focus on developing core competency in this needed area.

Figure 5.7: Process Improvement Work Breakdown Structure

In Figure 5.7, each activity is broken into related tasks. A task is defined as the smallest unit of work subject to management accountability. Once properly defined, tasks generate products, consume resources (people, time, and so on), and may be related to or constrain one another. For example, the support environment provides stakeholders with timely access to process resources stored in the process asset library via your Web site. Two tasks are involved in this activity: developing the Web site and hosting the process asset library (PAL) on it. As already mentioned, the PAL is the mechanism that the process group plans on using to make its processes, training materials, newsletter, and help desk support available via the firm's network to those who have authorized access. In addition, the PAL facilitates collaboration with working projects through specialized tools such as Microsoft's NetMeeting.

As you talk with your stakeholders about implementing these steps, the consultant offers the following suggestions relative to their accomplishment:

Establish expectations or management will set them for you. Work with your stakeholders to set realistic expectations for reaching Level 4. Otherwise, the expectations they set for you may not be achievable. Set reasonable but aggressive goals that you can live with. Use these goals to prioritize the tasks to be done by their perceived importance. As part of this effort, brief management, and get them to concur with the group's ratings. You will need everyone's support at one time or another as the initiative unfolds.

Do things that middle managers think are important. Respond to your middle managers by completing tasks that they feel are important to the success of their projects—typically tasks that will have a positive impact on their project's ability to deliver promised products on schedule and under budget. If you help these managers, they will help you. Make them look good, and they will become your strongest supporters. Fail to deliver, and failure will haunt you forever.

Do the easy things first. As you prioritize tasks, look for easy things that can be done quickly to generate positive effects. Completing them will help you display your initiative in a positive light. Perceptions are important. Middle managers view overhead activities such as yours negatively especially when they don't see anything of value coming from them. To counter this, promote your early successes. Make the people in the middle believe that you are a producer. Then they might be more apt to work with you when you ask them to pilot your processes on their next major upgrade.

Start with an operational concept. Develop operational concepts for use as you kick off your Level 4 initiative. A list of concepts is provided along with a brief summary in Table 5.2. These concepts should be couched in terms of the methods and tools that you will use to deploy needed new processes, pilot them, and transfer them into use throughout the organization. Most important, these concepts should provide management with feedback about your progress and whether or not you are achieving your numbers.

Then develop an actionable plan. Using these operational concepts and your WBS, develop a plan of action and milestones aimed at realizing the expectations you've set. Be practical: focus on the things that can be easily accomplished. Emphasize tasks in your plan that will move your Level 4 processes into project use in an ordered and systematic manner. Your challenge is to get busy people (e.g., the influence makers and the 20 percent responsible for 80 percent of the work) to use something new. That's why you need to budget to support the early adopters who agree to try the new processes.

Use available, free resources. Try to take full advantage of the resources you have access to and are paying for as part of overhead charges. In other words, tap the expertise that exists within your firm in areas where you may need specialized help (accounting, legal/licensing, and so on). The trick is finding out where and to whom to go for help. For example, find an expert who understands intellectual property rights to help you when dealing with software licensing issues. This person can help you cut through the legal mumbo jumbo and exercise the advantage you may have when negotiating licenses with vendors. Be careful to avoid"experts" who say it can't be done. Their job is to tell you how to get it done, not to try to deter you from innovating.

After listening to the consultant's suggestions, you continue with the steps illustrated in Figure 5.6.

Operational Concepts


Process development

Address the approach you will use to get to Level 4. You plan to exploit industry experience and best practices to expedite this job. You will hire an outside firm to help you, especially in statistical process control. This firm will play a major part in the assessment.


Get projects to use new processes. You will pilot the use of the processes to demonstrate their feasibility. You will also identify early adopters who will work with you as part of the pilot evaluation team to bridge the gap to widespread adoption.


Determine how to deploy processes once they are released for use. You will provide a process expert as liaison with the project. Education and training will be provided on a just-in-time basis.

Configuration management

Focus on maintaining integrity and version control of process group products. You will use existing methods and tools to do this job. Your steering group will be your change control board.

Quality assurance

Make sure processes are not released prematurely. You will enlist stakeholders to do work product inspections using the process in place for software development.

Distribution management

Address how processes will be distributed as they are released/updated. Your initial thoughts focus on providing access via a Web site.

User support

Answer users' questions, and provide users with timely support. You will staff your Web site with a part-time person. You will publish frequently asked questions and capture metrics on user satisfaction. In other words, you will run the site like a business.

Build Partnerships

Plan to collaborate with those people and projects that agree to support your effort. Deal only with people you know and trust. Ask these key people to serve on steering committees and working groups and to provide you with pilot projects. Ask advice, and seek their confidence. Recognize that you have to give something to the partnership. If all you do is take, what's in the relationship for others? Volunteer to spend money to help your partners succeed. Put people on the project to off-load them so that they can satisfy their obligations. Do things on a noninterference basis so that it won't seem like their success is in jeopardy.

Plan to Sell, Sell, and Sell

Recognize that those in positions of power change as people move on for whatever cause. Just as soon as you've educated one set of executives, another set comes in the door. Don't get too comfortable. Be vigilant, and look for dragons. An overhead budget the size of yours is always a target for takeover. Remember you are a success if you are viewed as one. In response, create the illusion that you are successful in all your actions, lectures, and written work. The saying that "success breeds success" is not just a platitude; it's reality. Projects will support you when it is thought you will deliver. Ride the white horse, and people will wave as you pass by. This discussion stresses the need for you to promote your successes, even when they are small ones. Recognize that the sum of a large number of small successes is often perceived as one large success.

After several iterations, you feel comfortable with your plan of action. You've also completed a top-level business case using productivity improvement to justify your expenditures. The consultant recommends that you review the plan in its draft form with your champion to make sure that it meets the mark and you haven't forgotten anything. Your champion is busy, but his second in command assures you that he will like it.

  • + Share This
  • 🔖 Save To Your Account

Related Resources

There are currently no related titles. Please check back later.