Print

Comment

Email

Digg

Del.icio.us

Reddit

Twitter

Bob and Shaw have to test a particular software application and functionality for a period of 5 days. The expectation set by the Project Manager is to execute predefined set of test scripts for some modules. They are also given 8 hours a day to complete the task.

Bob's deliverable after a week

• Task completed in 4 days by spending 10 hours at work every day
• 18 new bugs discovered in the software
• New test scenarios were added for the module under test since the existing suite did not provide complete coverage
• Leveraging on past experience, a root cause analysis of the bugs
• was performed. The results were compiled and used to create a handbook on how to avoid common bugs in the application
• Shaw's deliverable after a week
• Task completed in 4.5 days by spending 10 hours at work every day

In both cases it is obvious that the tasks were completed ahead of time. But how can the Project Manager quantify the performance of the Bob and Shaw?

• Bob and Shaw completed task before time.

• Bob's productivity is higher than Shaw during bug verification (assuming that all the bugs for them are of similar nature).

• Bob spends more time in providing additional benefits for the project (client) and for the organization

Apart from the bug verification effort of Bob and Shaw, all other activities of Bob are not quantifiable.

Lack of understanding of software project metrics is one of the key issues faced by project managers and is one of the reasons for poor project management. It is important to define which metrics to capture and how to apply them in the project. This is the basis of metrics design. For instance, if the release of the product is the driver for the project success, then a project plan should have well defined intermediate milestones placed in sync with the critical phases of the project and tracked. If a metric is designed to the success criteria for every intermediate milestone, it is possible to control the project better and steer it in right direction.

It is also found that gathering metrics for the sake of measurement is ineffectual. Every project manager strives to deliver a zero-defect product on schedule. The end-result is dependent on a multitude of factors, with efficiency, skill-set and productivity of resources playing an important role. Both quality and productivity directly depends on the performance of the project members. That "You cannot control what you cannot measure" is true. It is a known fact that by measuring the performance of every activity and computing the metrics on a frequent interval, it is possible to control the project effectively. Data points are required to control the budget, schedule, cost, effort, quality, productivity, client expectation, senior  management expectation and team performance.

One way to do this is to use a Weighted Metric Value-added (WMV) methodology that deploys quantitative project data and performance measures from day-to-day project activities for successful software project management. This method is based on task metrics covering all aspects of project activities, and assigning a score for every activity. The scores are used to manage the project. The objective is to help a project manager control and steer the project in the right direction using facts and data, rather than misleading optimism.

The WMV approach can predict project performance, team performance and the impact of various management decisions. An attempt is made to emphasize the importance of adopting quantitative techniques for managerial decision making during a software development lifecycle. Using this approach one can measure and monitor the trends in the project quality, productivity, benefits to the organization and the client as illustrated in this paper. Further, this methodology helps to make better decisions on critical aspects such as resource planning, quality of the deliverable, effort tracking or slippage.

Task-based activities and value-added activities
Fundamentally, all software metrics are various permutations and combinations of the following parameters

• Time-Schedule adherence

• People-Performance of the team

• Quality-Delivered defects, meeting client expectations, process compliance

• Scope-Managing requirements churn

• Cost-Managing the budget

• Risk-Managing the risks

These metrics are derived from the tasks performed by the team. Value-added activity encompasses a gamut of additional services provided with the intention of building a long-term relationship with a client. The cornerstone of this relationship is the confidence that the client vests in the organization. Hence it is important to capture both the task activities and the value-added activities from each project team member to the client on a regular basis and use this data for planning future strategies. Value-added activities help in understanding the potential of each member in the team and thereby it is important for the Project Manager to quantify these activities in the project.

Some of the value-added, such as completing task ahead of time, can be measured easily, but there are other activities like creating short FAQ's for newcomers or creating a trainers' checklist which go unnoticed in the absence of any tracking and monitoring mechanism. Thus it is very critical for a new evaluation process to be incorporated for quantifying and capturing these day-to-day contributions of the team members. A sample representation for assigning score based on the metrics and activity is explained in 'Sample Project CODETEST information' (see table in previous page).

How to track and collect data?
The foremost requirement for executing the project is to collect and track data on schedule, effort, cost, quality and performance of the team. Effort, quality and cost can be periodically tracked using standard in-house project management tools. Though there is no standard guiding principle on the frequency of tracking project data, daily tracking is recommended for best results. Daily status tracking can be very effective in controlling projects and is widely practiced in CMM© Level 5 organizations.

Daily Activity Report: This has the activities done on the project and the corresponding time spent in each activity.
Daily Status Report: Some important data that can be collected from the daily status report include:

• task accomplished and pending

• task planned for the next day

• task planned for the week

• list of issues and status of the bugs

• soft and hard target data (can be done on weekly basis)

Lessons Learned
The Weighted Metric Value-added approach was found to be useful to measure the performance of the project team members. It provides a yardstick and a good handle on the consistency of performance and ability to the project manager to work on the areas of improvement for self and the team. It also clearly demarcates value-added activities from the mainstream project activities. Clearly, the approach could go a long way to help project managers conduct effective performance appraisals. However, the negative usage of metrics may back fire on projects. The data collected must be used to improve team performance, and not to focus on faults alone, minimizing negative team dynamics, thereby offsetting data manipulation.

The WMV approach may have subjectivity in the beginning of the project, but it can be fine-tuned and refined as more and more data points are tracked.

Kiran Kumar Marri, project manager, Infosys Technologies, Bangalore

Next Page :

Classification of Tasks

Page(s)   1  2  3