To be effective,
performance
measures must
be properly
linked to the goal
of value creation.

In the conceptual focusing matrix (see Figure 3), the top-right-hand quadrant is designated “pearls” since it contains the most valuable candidates for selection. The three other quadrants are designated “oysters” (valuable yet hard to crack), “low hanging fruit” ( easily accomplished but less valuable), and “white elephant” (to be avoided). This terminology facilitates communication among managers.

Top management takes the final strategic-gating decision, deciding to include in the realization portfolio oysters if it views them as having strategic importance beyond the value-creation estimate. Likewise, top management could consider inclusion of low-hanging fruit if it considers it appropriate for achieving quick successes.

An outside observer might doubt whether division heads, being eager to promote their software-solutions requests, would deliberately exaggerate estimates of value creation. To obviate this risk, every request for a software solution must be backed by a business plan (or at least a mini-business plan). In addition, the division heads must commit to the anticipated value creation and consequently include the derived cash-flow estimates in their revenue targets. A follow-up mechanism concerning how to achieve these value-creation targets should be linked to the organization’s key performance indicators and incentive systems.

Waste Can Be Avoided

Research shows at least 50% of IT developers’ time is wasted11 due to the following reasons:

˲ Rework due to incomplete or poorly defined needs and requirements (“incomplete kit”);

˲ Rework due to frequent changes in requirements and scope up to the final delivery stages; most such changes are not “must have” but only “nice to have”;

˲ Software solutions developed and delivered but eventually not used ( happens all too often);

˲ Over-specification of requirements to include functionality and features seldom or never applied; and

˲Having too many activities assigned to individual developers, leading to wasteful context switching among activities (bad multitasking).

Waste cannot be totally eliminated but can be reduced significantly through seven simple managerial practices, or remedies:

The complete-kit concept.e To avoid waste of IT resources due to rework, the organization, as a whole, should adopt and implement the complete-kit concept. 10, 11 Projects should not be approved for development if their business rationale, requirements, business plan, or statement of work are only partial or vague. Likewise, tasks should not be assigned to individual developers if the requirements/specifications/ design are incomplete or fuzzy. The content of the complete-kit definitions for the requirements and specifications is listed jointly by the requesting divisions and IT personnel.

Since requesting divisions must provide complete kits of requirements and specifications, they must understand their needs and the required software solutions. This would minimize the extent of requirements changes with resulting rework and the number of projects developed and delivered but eventually abandoned. Moreover, implementation of the complete-kit concept usually leads to improved communication and collaboration between business divisions and the IT division.

Also needed are definitions of complete kits for tasks and activities performed along a project’s life cycle; for instance, the system-analysis files given to programmers by their managers must contain a complete kit of information to enable them to do their jobs properly. Likewise, files given to testers must contain the complete kit of information to enable testers to complete their jobs properly.

Eliminating over-requirements, over-specifications, overdesign. Scrutiny of software-solution requirements usually reveals a high degree of over-requirements. Significant portions of functionality and features required for a software solution are nice to have rather than must have. In many cases IT professionals tend to introduce over-specification and overdesign to be on the safe side in e The complete-kit for a task is the list of all items required to complete the task without interruption.