Real Process Improvement Using the CMMI

How you go about implementing process improvement in your organization will be determined to a great extent by the prevailing management and cultural attitudes toward good, fast, or cheap. Every day in every software or system organization in the world, people choose between good, fast, or cheap. These choices are so ingrained, so institutionalized as to almost be the DNA of corporate management. Even in the rare situation in which an individual is conscious that she is making this decision, she won t admit it. More often than not, managers publicly deny making such choices by giving lip service to one choice or the other depending on the audience. To the employees and customers, quality is our top priority (aka, good). To the middle and senior managers and to marketing, we ll meet the deadline (aka, fast). To the executives and board members , we ll give you the best value, ROA, or ROA (aka, cheap).

Most traditional approaches to process improvement that I ve experienced or observed achieve only one of these goals: fast. Fast maturity levels also seem to be the favored hot product being sold by many process and CMMI consultants . [25] There are essentially six approaches which heavily influence how good, how fast, or how inexpensive your process improvement project will be. These basic approaches are:

  1. Leveraging native standards and practices versus starting from zero (read Natural Process Improvement through Weeding and Nurturing in this chapter).

  2. Aligning process improvement with the organization s business (read Business and CMMI Alignment in this chapter).

  3. Learning vicariously versus online learning (read Organizational Learning and Process Improvement in this chapter).

  4. Focusing on business problems and goals versus focusing on the maturity levels.

  5. Implementing a work product-based approach versus procedure-based approach (read the section titled the same in this chapter).

  6. Pursuing a meta-feature implementation versus PA or practice implementation (read Integrated versus Vertical Approaches to Process Improvement in this chapter).

The following subsections will either help you identify the nature of the process improvement approach that may already be underway in your organization or help you decide which approach is best for your organization. The approaches are then described individually in subsequent sections.

Fast and Expensive Maturity Levels; Not Good Process Improvement

Organizations which experience rapid implementation of CMMI-based process improvement, such as the organizations whose maturity level achievements mark the quick end outliers in SEI s Maturity Profile, [26] are usually after one thing and one thing only: maturity levels (see Figure 4.6). Fast, expensive process improvement is prevalent in organizations that perceive maturity levels as a preeminent marketing tool for getting new business. The segments of industry in this crowd include IT outsourcing organizations trying to out-do competitors in the levels race and non “U.S. companies seeking a piece of the U.S. software and systems market. In these environments, the race to achieve maturity levels resembles a full-on death- march project.

Figure 4.6: Fast and Expensive Maturity Levels; Not Good Process Improvement

The quality of the results, e.g., the business benefits from process improvement, may be good or poor, but it s likely that no one in the organization will know due to the fact that the only thing tracked, measured, and reported is the achievement of maturity levels.

Approaches Employed

Organizations racing to a fast, expensive maturity level typically pursue this combination of approaches:

Additional Characteristics

Some other general characteristics of fast, but not good or cheap process improvement include:

The reason these types of process improvement efforts are so expensive is because relatively lots of money and effort are expended on developing standard policies and procedures without consideration of reusing or making incremental, lower cost improvements to existing engineering and management practices. Also, by nature, such efforts coincide with a management culture that can best be described as centralized command and control. In these types of organizations, a one- size -fits-all mentality often pervades and standardization is sought for its own sake. So, at the level of implementation, much effort is expended by project managers and engineers trying to find ways to force-fit the centralized standard procedures to their projects. If the standards and procedures are overly bureaucratic, effort will be wasted creating project documentation that provides no measurable value to the project, the customer, or the company.

The fast and expensive maturity levels problem is worse than anywhere else in the IT and system engineering outsourcing business. In one such organization in which I was a process improvement manager, the people negotiating new contracts committed the applications personnel to an impossible , no-way-out situation. In their game of outdoing the competition s bids for contracts, the marketers wrote into the contract these two contractual (paraphrased) obligations:

The organization will be certified (sic) at SEI (sic) Level 3 within 24 months after the cut-over (of application support to the outsourced vendor).

Many pages later in the contract, you find among the contractual commitments:

At no cost to the client, the vendor will provide the client with detailed estimates and schedules for projects within three days of the approved service request.

What s so wrong with these two contract points? Think about the behaviors that will be driven by the fact that the client cannot be billed for a critical activity of project planning. If project planning, or at least some of it, has to be free, won t the outsource company s project managers have incentive to produce plans as fast and as cheap as possible? Think about this situation in the context of the Project Planning PA and GP 2.2 of CMMI. Let s face it, no one ” not even the best project manager ” in this kind of environment is going to establish and maintain realistic plans. Why should they? Where s the incentive?

Fast and Cheap Maturity Levels; Not Good Process Improvement

Some organizations also quickly attain CMMI maturity levels without overspending (see Figure 4.7). At the risk of giving you any not-so-bright ideas, the approaches employed and the general characteristics of such approaches are described in this section.

Figure 4.7: Fast and Cheap Maturity Levels; Not Good Process Improvement

Approaches Employed

Organizations racing to a maturity level sometimes do so relatively inexpensively. The quality of the process improvement may be good or poor, but no one knows because it s not measured.

Additional Characteristics

Organizations that pursue fast and cheap maturity levels, but not good process improvement also exhibit many of the same characteristics as organizations that go for fast, expensive maturity levels. Some additional general characteristics include:

Balanced Process Improvement; Fast Enough, Reasonably Priced, and Good Enough

Whereas achieving the best quality at the lowest price in the shortest time is probably just wishful thinking in any business endeavor, the reality is that you don t have to choose the polarized extremes of any of these goals (see Figure 4.8). In model-based systems process improvement, you can achieve a quality that provides measurable benefits to the organization and its customers in a timeframe that is competitive and at a cost that doesn t kill the organization.

Figure 4.8: Balanced Process Improvement: Fast Enough, Reasonably Priced, and Good Enough

Approaches Employed

The combinations of strategies to take to achieve balanced process improvement are:

Characteristics

In organizations which strive to achieve measurable business benefits using CMMI for process improvement, you will often see these characteristics:

[25] An unsolicited advertisement from Castilon Consulting LLC dated July 15, 2003 sells, Level 0 to 3 in less than 6 months. The brochure cites a case study in which Castilon Consulting helped an IT organization rapidly achieve CMM Level 3. In the Benefits section of the advertisement, Castilon identifies 13 benefits that can be realized from their approach. However, nowhere in the entire brochure do they mention anything about their clients achieving even a single business improvement (e.g., ROI, ROA, productivity, product or service quality, customer satisfaction, employee satisfaction) from the Castilon approach. Apparently, the rapid achievement of a CMM maturity level is the benefit. They re probably doing okay since there seems to be quite the market for maturity levels.

[26] The Maturity Profile is a report periodically published by the Software Engineering Institute (SEI). This document provides information and statistics from the results of CMM and CMMI appraisals reported to SEI. For more information, go to www.sei.cmu.edu.

Категории