Words and figures at work (making estimations)

Plans are based on estimations. Like gambling, the math can be useful to make a decent living of it.  However, figures shall be shipped with a pinch of creativity (forget the fads set by financial market).

In the specific situation, the art of communication is applied to supply the developers, who remain responsible - within the team – for the estimations’ process (Martin Fowler ). The project manager shall set everything so that they receive:

• the requirements (technical or external / “commercial”) that form the scope
• the constraints (time and effort. It also should include company’s policies and strategies).

• Either the focus is on the big picture or on the single object; developers shall receive clear instructions about the business scope.
• The project manager role consists in the action of mentoring the figures that have to be framed into the business case / plan, then to sustain those figures in front of the stakeholders.

The sense of purpose (proportion)

There are two different kinds of sieves to be used by project manager / team leader to evaluate the level of details:

• Time proximity. Nearer the task to be estimated, narrower the matter to be discussed. Therefore, the nature of details will change.
• Functions and skills. Some technical objects (e.g. databases) can be designed in two stages. Starting as macro (business) entities, then into specific details (e.g. clustering, store procedures, tuning,).  Therefore, different skills are demanded to carry through this kind of estimation.

Splitting the main job into different phases and different people could offer a good starting point. Each group decision will be presented and discussed with people outside the former cluster. The idea is about challenging the proposals in order to gain a broader support. The costs (duration multiplied for each participant) of these meeting shall be “paid” by the accuracy of the results.

The project manager must have the competence either to (broadly) understand the pros and cons of the choices or delegate to the right person(s) the following tasks:

1. Architecture
2. COTS (whenever is it feasible to use them)
3. Quality
4. The correct mindset (within the producers’ team) to deal with subsequent requests from “external” stakeholders
5. Escalating issues procedures
6. Tool for process alignement
7. Configuration & Change Management – project’s assets
8. Tools and procedures to deal with tests
9. Organizational methods within the producer team
10. Strategies for implementing and maintaining the project’s outcome.

This is also a good opportunity to carry out the routinely risk assessment routines.

Updating event proximity

Has change something about the timing of the cause spotted during the Risk Analisys

Reviewing triggers criteria

Are the values considered acceptable still valid for this task?

Risk attitude

Is the team/operator confidence’s level sustained by valid elements (i.e. previous success?)

Conclusion

There are some very good books and articles about this essential task Software Estimation, on the net you will also find nice articles that dispute the concept itself.

When words are scarce they are seldom spent in vain