There are number of different approaches to cost estimation. The challenge is to find a blend of techniques that provides you with a “rough order of magnitude” very quickly and as more data becomes available, the estimate can be adjusted accordingly.

But before we go there, what exactly is an estimate when we are talking about IT projects? The Oxford dictionary defines it as

“a judgement that you make without having the exact details or figures about the size, amount, cost, etc. of something”

When IT projects are funded, the Cost Estimates quickly become Cost Commitments.

The following is a list of cost estimation methods:

  • Expert judgement

“Expert judgement” has many challenges related to bias but there are techniques available to make it a very quick and fairly accurate method of doing “rough order of magnitude” cost estimates. This assumes that the “experts” have been with the organisation for some time with domain knowledge in the area of the estimate.

  • Top down

“Top down” is similar to expert judgement based on a high level view of the project requirements. It is useful to try and define the range that the project is likely to be within i.e. probably no bigger than “X” and no smaller than “Y” and with a duration of approximately “Z” months.

  • Analogy based

“Analogy based” estimates can be very good if the organisation has good data from previous projects. We often find that data from previous projects is kept but not neccesarily in a format that can be used for comparisons. If there is good project data available then the current estimate can be compared to similar previous projects.

  • Bottom up

“Bottom Up” estimates are the most accurate but require that all aspects of the detailed requirements have been completed. This means that a lot of design and analysis work has been completed without knowing if the project will be approved.

  • Knowledge based

“Knowledge based” estimates are used where there is domain knowledge data that can be used to understand role, effort ad duration for the elements. Different element volumes can be applied to get the estimate.

  • Parametric

“Parametric based” estimates use relationships between elements to develop the estimate. The SEER suite of tools uses a combination of knowledge bases and parameters to quickly develop comprehensive cost models.