Art of Project Estimating

Since there is no such thing as an exact estimate, what is the best way to forecast costs on the project.

e do project estimating for one of two reasons - we want to determine the costs to establish a budget or we want to determine if the project can be accomplished within a pre-established budget.

Project costs are time-phased. We are concerned not only with how much the costs will be, but when the expenditures will be made. When you're working a long running project of a year or more that may require outside funding, your project estimate should show approximately when the various funds will be needed. Some purchases will occur at the beginning of the project while other costs, like labor, will be spread over the duration of the project.

Types of cost - Costs may be incurred for materials, services, and overhead. They may be fixed, occuring one time like purchased materials.  They may be fixed and recurring like equipment rental or contracted services. They may be variable like labor or consumables on the job. Overhead, also known as burden or general and administrative costs, generally represent support and management costs.

Prerequisites to project estimating - Before you can create an project estimate you need to know what has to be done. That means you will have an accurate list of deliverables, and the associated bills of material and work breakdowns associated with those deliverables. Additionally, you should have a good idea of the resource requirements -- what people or contracted services you'll require. Once you have that information in hand, you're ready to begin defining the fixed and variable costs associated with each work package and deliverable.

Bottom up approach - The most accurate way to estimate costs is with a bottom up method. With this approach you determine costs at the lowest levels of detail and then roll or sum them up to get the values for higher levels including the entire project. This approach requires well-developed bills of materials and work breakdowns with a sufficient level of detail that cost and effort can be accurately estimated.

Estimate as a range of numbers - There is no such thing as an exact project estimate, no matter how often one might be requested.  Estimates should always be presented as falling between a range of numbers that represent a best and worst case.  Project estimating should never result in a single value.  If you determine the best and worst numbers at the lowest levels and roll up both sets of numbers you'll have a reasonable range of values for the project costs. The best case should reflect the most realistic probable costs. The worst case should include all contingency costs arising from risk assessment and other issues. Overhead costs are then added to the estimated values, most often as a percentage of the estimated values.

Getting estimates from others - When you are asking others to estimate costs in their area of expertise, be sure to ask for the estimate as a range of numbers, not a single value. It may help if you suggest that they use the best case and worst case method discussed earlier. You should also ask for a list of the factors that would drive the costs from one point in the range to another.

Setting a cost baseline - when you have the project estimate in hand and have presented it to appropriate parties, get commitment for the budget and use that as a cost baseline. As the project moves forward you can track expenditures against the cost baseline to verify the accuracy of your project estimating efforts and to make sure that you are staying within budget.

- Jake Alexander

 next article: Keeping a Project Management Journal