Estimating is the activity of predicting what a piece of work will require in terms of time, resource and cost. This can range from a high level estimate of a project in a programme to detailed estimating of individual activities in a work package.
Almost all estimating goes through two stages: top down and bottom up.
In the early stages of the project or programme approximate estimates are needed based upon limited information. As the work becomes increasingly well-defined greater accuracy can be achieved, i.e. definition and the estimate are developed from the top down.
Once the detail has been established small chunks of work can be more accurately estimated. Individual activity estimates can be consolidated to develop more accurate total costs and timescales, i.e. from the bottom up.
With differing application within the top-down and bottom-up contexts, there are four fundamental approaches to estimating:
- Comparative (also known as analogous)
These terms are not mutually exclusive. For instance, a particular method may combine parametric and subjective approaches; another may be a combination of comparative and analytical. Estimating methods are as diverse as the range of organisations who undertake projects and programmes.
The approach or combination of approaches used will depend upon the technical nature of the project and the life cycle phase.
To illustrate the different approaches, consider the problem of deciding how much to insure your home for in the event that it needed rebuilding.
- The parametric approach is the one generally taken by insurance companies. The cost of rebuilding is based upon certain parameters such as:
- Number of bedrooms
- Garage/No garage
Number of floors
Depending upon the parameters above a series of formulae would come up with a rebuilding cost. This example of a parametric approach is clearly early on in the top down stage.
In the analogous (also called comparative) approach you would ask people who had recently had homes built, how much the work cost. You would adjust the figures they gave you to account for the differences between your house and theirs. To do this comparison you need more detailed information about how your house is different from those used for comparison and so is later in the top down stage.
The analytical approach would mean calculating building costs from first principles and is what a builder would do if provided with a full set of plans for a house and asked for a price. This would be built up from the quantities of materials, productivity figures for different trades and hire costs of machinery. This is very much a bottom up process and would be based on a product or work breakdown structure.
Not all projects lend themselves to physical measurements of the work to be done. In many instances the approach must be a subjective one typified by the question: “Based on your experience, how long will it take to do this activity and what will it cost?”
Estimating forms the basis of promises to stakeholders and inaccurate estimates can be the source of much conflict. If everything could be predicted with 100% accuracy there would hardly be any need for project and programme management but that is clearly not the case.
Many approaches to scheduling have been developed to address the inherent uncertainty of estimates. Monte Carlo, PERT and critical chain techniques apply different combinations of statistical and psychological approaches to provide more pragmatic schedules.