- Check the history about how similar projects were estimated.
- Partial WBS is also an option.
- Sometimes Expert Opinion is the best answer.
- It should be based on the Solution Architecture.
- It should take bottom-up approach.
- Divide and rule. Make it modular.
- (Development) Language does matter.
- Offshore can reduce costs but not all the times.
- Make sure that what-if scenarios are identified
- Feature and price breakdown are necessary evils.
- Sometimes, phased estimation saves the project.
- Some of the following methods are good estimation-technique-candidates:
- Parametric,
- Educated assumptions,
- Delphi method,
- Dependencies identification,
- Previous history
- Task breakdown
- Assumption is the chair on which estimation sits sometimes.
- Distrust can undermine estimates.
- Padded estimate generate distrust
- Estimating techniques can help manage the risks
- Guesstimations can screw the projects.
- Use Case Metrics is also a possibility.
- Ranged estimations are the answer when you do not have precise information.
- Three Point Estimation technique is considered a reliable one.
- Ultimate destination of an estimate is Budget.
- Not all management knows the distinction between estimation and budget.
- Make sure to add estimates for User Acceptance Test (UAT) – it will save the project from being a failure.
- Variations with the actual and Ballpark estimations can be up to 300%
- Variations with the actual and Rough estimations can be up to 50% to 100%
- Variations with the actual and Fair estimations can be up to 25% to 50%
- Have you considered contingency plans in the estimate you just created?
- Those who do the actual work should be an integral part of estimation.
- Estimate of all 4 variables is must: time, people, and resources and the risks
- A good estimate does not need fancy tools – Ms-Excel or a pen and paper would suffice if it is executed nicely.
See Also: