One of the areas that has always amazed me with respect to Alliance Partners and their system integration efforts is their ability to estimate and ultimately bid the project cost. It seems almost an art (perhaps a black art) as it is a science, but it is critical to the success of the business.
It starts with gathering good requirements which is an art in and of itself. In a previous post, we discussed a ‘needs-based’ approach and the type of requirements that you want to gather. We also talked about probing techniques and how to deal with conerns.
Once you have a good set of project requirements, you still face the challenge of estimating the cost of the project. Hopefully, you have a good idea of the hardware requirements, but labor usually makes up the bulk of the total cost. Ideally, you can break down the project into a set of tasks to develop a system that meets the requirements. Then for each task you can estimate the effort, duration, and cost.
There are several methods for estimating the labor cost of each task:
- SWAG (Not recommended) – Scientific Wild Ass Guess
- Expert – ask an expert based on their experience
- Delphi – Ask a group to brainstorm and build consensus
- Comparative – Based on history or similar task
- Weighted average – a combination of estimates (e.g. Optimistic + 4*Likely + Pessimistic) / 6)
Obviously, depending on the size of the project, you’ll need to decide how much time you want to invest in this process. But, estimating project costs is critical to your profitability.
Labor Cost Variables
Other factors to consider in estimating projects.
- Work interruption factor – no one can work uninterrupted. For instance, meetings. Such breaks in the thought process and work effort will inevitably impact effectiveness. Idea: consider limiting meetings and e-mails to certain parts of the day.
- Part-time effect – Working on multiple projects can impact effectiveness. The developer must ramp up or down from one project to the next.
- Skill factor – Obviously, a novice may take longer to complete a task than an expert.