Practical Solutions for your Software Development Challenges

Vol. 4, No.2 August 1997

An Objective View

Mary and I have been performing SEI process assessments since 1989 and have been authorized assessors since 1990/1991. Our role on an assessment team used to be primarily as therapists, pointing out "stinky fish" issues that no one else was willing to talk about. Over the last five years, many companies have significantly matured, both in process and people skills. While there are instances where we point out "stinky fish," our role, in addition to leading the assessment, has become primarily that of answering Capability Maturity Model (CMM) interpretation issues (much of which is in the CMM examples text).

When you perform an assessment, particularly to verify that your organization is level N, ensure that you have someone from another division or company that can be your objective eye. Without an objective eye, you run the risk of the organization rationalizing the current behaviors to be whatever the goal is.

Rationalizing is not bad, it is human nature. It allows us to see things consistent with our beliefs. If we didn't rationalize, we might go insane due to an overwhelming set of choices in life. In an assessment, rationalization can become a problem, since it can filter out issues that need to be addressed,or strengths that are taken for granted.

With the need for more companies to be SEI level N, we are seeing more organizations declare how they were assessed and who was involved. This is a positive sign and maintains the assessment standard. No longer is it credible to skim through the CMM and claim one is solidly level 3.

--- Neil Potter

Top of newsletter


Planning Improvement


The starting point of an effective process improvement program is to conduct a formal or informal process assessment to determine the strengths and weaknesses of the organization. After the problem identification phase, a high-level improvement plan is created. This is followed by a detailed tactical action plan. This article describes the steps to create a high-level, or strategic plan.

The overall flow of the strategic planning process is:

  • Select a business goal on which to focus
  • Select assessment findings (or problem areas) that support the business goal
  • Develop subgoals for each problem area
  • Prepare for tactical planning

In detail, the steps to develop a strategic plan are:

1. State the business goals of the organization being improved.

The business goals are used to determine which assessment finding to address first, and to keep the team focused on a clear result. If the improvement effort does not help the business, it may fail due to a lack of management support. Using the business goals makes the tie clear.

2. Select one business goal as a focus.

3. Select a strategic planning team.

Look at the business goal and problems you are trying to solve and choose people that have a vested interest. Include people, when appropriate, from management, development and marketing.

4. For the selected business goal:

(a) If a process assessment has been conducted, determine which assessment findings (or parts of an assessment finding) would impact the goal the greatest.

(b) If a process assessment has not been conducted, brainstorm and discuss the problems that are preventing you from achieving the goal, and set priorities.

5. For each weakness (assessment finding or problem):

(a) Each member of the planning team gives his/her interpretation of the weakness. The team validates its understanding and discusses and resolves any differences.

(b) Determine subgoals (over the next 18-24 months):

How should the organization look or behave when this weakness has been addressed? These subgoals (established for each weakness) should be in line with the overall business goal.

The creation of an image of how you want the organization to look after an assessment is the most important part of strategic planning. Initially you may come up with images that seem unbelievable or ones that you don't feel you have any control over. These will be refined in step "d."

A good analogy is road building. The business goal describes where the road ends. The subgoals, based on problem statements, state the potholes that need to be filled in order to achieve the business goal.

(c) State why this subgoal is important to achieve (the benefits).

It is important for improvement goals to be compelling. For each subgoal, state why the organization wants this to be achieved. This is the motive. If you cannot develop a compelling argument for achieving a subgoal, you may have found an area not to waste time on for now.

(d) Check to make sure each subgoal is SMART: Simply stated and specific, Measurable, stated As-if-now, Reasonable, believable (within your control or influence), Timed (with a date) and toward what you want. See page one of this newsletter for an explanation of SMART goal setting.

6. Set priorities for each subgoal statement.

It is important to focus on a few areas and not try to solve everything at once. Priorities can be set by determining the relative benefit and cost of each goal.

  • On a 1-10 scale, rate the relative benefit the subgoal would have for the organization (1 = low, 5 = medium, 10 = high)
  • On a 1-10 scale rate the relative cost of the subgoal to implement (1 = low, 5 = medium, 10 = high). "Cost" can be money or effort.
  • Determine the priority (benefit / cost).
  • Any interdependencies? Categorize into phases, 1st, 2nd, 3rd.

Look to see if there are any natural sequences for the subgoals and then sort the complete list by phase (use three phases for simplicity). For example, when working on planning, a subgoal in phase one may be to identify a project planning technique and pilot it. A subgoal in phase two may be to plan one real project and collect data. A subgoal in phase three would be to deploy the technique to all project teams. Start work on those subgoals in phase 1 with the highest priority.

7. Determine when the goal and progress should be reviewed.

If the goals are not frequently reviewed, they are likely to get out of sight and out of mind, and action towards their accomplishment will fade. Frequent review will maintain a burning desire and consciousness for the goal so that the organization stays focused.

8. Do a preliminary brainstorm of tactical ideas.

At this stage it is useful to brainstorm ideas for the action plan. Not all of the ideas will be used. If you have a process goal of the CMM or ISO9001, pick out items that help address the problems and achieve the subgoals. The quickest way to achieve any process model or standard is to find some practical use for each element. The selection criteria for where to start looking are the business goals and problem areas of the organization.

Example - A Portion of a Strategic Plan


Business goal = Time to market for product development is six months on December 15, 1998




(Antithesis of
Problem Statement

Why do we want to achieve this subgoal?

Relative benefit of subgoal (1-10)

Relative cost of subgoal (1-10)

Priority (Benefit/


Planning is
ad hoc

Planning is structured,
i.e. a process is used. 

Better planning will reduce rework, costly surprises that slow delivery. 





Data exists to predict delivery. 

Data will increase credibility and allow better scoping. 





New project leads are trained in planning.

Training will help avoid mistakes and save time.





Top of newsletter

Writing (SMART) Achievable Goals


Goals that are well written are more likely to be achieved. When writing goals, make sure that they are SMART*: Simply stated and specific, Measurable,stated As-if-now, Reasonable, believable (within your control or influence), Timed (with a date) and toward what you want.

Simply stated and specific

When you call 911 with a report, specific instructions enable care providers to respond quickly. If you merely told them that there was a car wreck in the south, near a grocery, you would confuse the emergency staff . Maybe they would roam the south of the city, or call every grocery store inquiring about any nearby car wrecks.

Similarly, with improvement, vague goals are harder and slower to reach. Developing clear goals takes time, effort and refinement.


There must be an objective way to verify the achievement of the goals. Otherwise the team would have no way to tell that it had actually reached them.


When goals are stated as if they have been achieved, it helps us get beyond merely wishful thinking. When goals are written in the future tense it is easy to keep them mentally in the future. For example, with the goal "I will exercise more," I can consider the goal reached even if no action is taken, because I can sit in my chair and tell myself that "I will exercise more." Notice that because of the future tense, the statement is true today. If, on the other hand, I say "I exercise daily and I exercise aerobically three times a week for a minimum of 20 minutes," the brain can't ignore it. It tries to make the statement true.


To achieve goals, it is important that they are within our control or influence. Otherwise we can discount the goals as impossible and we're likely to procrastinate.

Timed and toward what you want

Goals eventually need to have a date in order to be effective, and need to be stated in the positive. That is, the goal should state what you want, not what you don't want.

Examples of SMART goals are:

  • Defects reported in the field have been reduced by 20% on July 1, 2000.
  • Planning predictability is 10% more accurate (compared with 1996) on December 31, measured by calendar milestone completion and effort required.
  • Productivity for coding (lines of code per 40 hours) has increased 10% each year.
  • SEI level 3 on July 1, 2010, measured by an external assessor.
  • 50% deployment of inspection process on Dec 31, 1998, as measured by process assessment.

Once you have written your goals, check them for clarity and specificity. Does the whole team or organization know what they mean, or do you need to refine them? If goals remain ambiguous, it is less likely they will be achieved.

* "Creating Your Future," Audio Cassette series, Tad James 1992.

Top of newsletter


Quality Software Management, Volume 4, Anticipating Change

Gerald Weinberg has spent three volumes telling us, in detail, what to change in order to improve. Now he gets to my favorite part: how to set up the right environment so that the necessary changes can occur. He explains various change models, ideas for change agents as artists, planning issues, and ideas about what needs to change and what needs to be preserved.

I like it best when he talks about setting up the cultural environment needed to make change successful, especially ideas for Change Artistry. As one would expect, he applies ideas from The Satir change model and lots of other tricks pulled from up his sleeve.

Here are some gems about change artists from my favorite chapter:

  • "They [change artists] dissuade people from attempting to short-circuit the change process with magic."
  • "[They] listen actively, listen honestly and clearly, set clear boundaries, allow others to struggle with their own chaos, interact with love and respect, and activate possibilities in each person."

He gives us useful planning information at a very practical level, including many of the "icky" issues others avoid, such as discussing lack of trust and addressing hidden agendas.

Overall this is an enjoyable book, full of many strong opinions, anecdotal comments, and practical hints about cultural issues. Moreover, there is still enough of what to change to keep the "techies" happy.

--- Mary Sakry

Weinberg, Gerald, M., "Quality Software Management, Volume 4 Anticipating Change" Dorset House, New York, NY, 1997.

Top of newsletter