The fourth process group in project time management is activity duration estimating. This is the essence of the entire time management process. The preceding processes have defined the what, when and who for each activity: this is the process that determines how long.
There are eight inputs to this process:
These inputs have all been described above, as they are used as inputs to the previous process groups. Ideally, the estimating of an activity duration should be carried out by the person doing the work. The role of the project manager is to select the most appropriate estimating tools and techniques and to ensure access to any existing historical or organizational information to assist the estimating process.
Tools and techniques
There are five tools and techniques for the project manager to use when estimating the activity durations:
Analogous estimating, sometimes called top-down estimating, uses the values of parameters, such as scope, cost, budget and duration from a previous, similar activity as the basis for estimating the same parameter or measure for a future activity. It is frequently used to estimate a parameter when there is a limited amount of detailed information about the project (e.g. estimating in the early phases). Analogous estimating is a form of expert judgement. It is most reliable when the previous activities are similar in fact and not just appearance, and the project team members preparing the estimates have been involved with the previous work.
Parametric estimating allows the project manager to quantify the activity durations by using a simple (often linear) assumption to relate the amount of work to the productivity rate. The productivity rate can be estimated using historical information or expert judgement. Examples of productivity rates include lines of code per hour, metres of tarmac laid per day, bricks laid per hour, or components machined per week. Johnson famously advised Boswell not to turn his land into an orchard on the basis of a parametric estimate: 'we compute in England a park wall at a thousand pounds per mile ... for a hundred pounds you could have forty-four square yards which is very little'.
The three-point estimate technique provides an average estimate of the time expected for an activity to be completed. This average is calculated from the most optimistic duration of the activity (O), the most likely (M) and the most pessimistic (P). So, for example, if the optimistic time for a component to be manufactured is 5 days, the pessimistic duration is 12 days and the most likely is 7, the average of these three estimates is 8 days and so 8 days will be used for the duration estimate.
The PERT (Programme Evaluation and Review Technique) estimating technique uses a weighted average to model the duration probability distribution, with M four times more likely to occur than O or P. The PERT distribution is usually normalized by dividing through by 6. For the example given above, the value of the PERT distribution is 7.5 ((5 +(4 x7)+ 12)/6). The mean of the PERT distribution is M. The variance of the PERT distribution (the expected value of the square of the difference between a given value and the mean) is therefore given by ((P O)/6)2 and the standard deviation of the PERT distribution (the absolute square root of the variance) is (P O)/6. So in the above example, the variance is 49/36 and the standard deviation is 7/6.
The final technique, reserve analysis, examines the risks to the project schedule and builds in buffers or time reserves. Building time reserves into a project should not be confused with padding the project. Padding is a term often used to describe extra slack or waiting time that has been built into a project to cover a lack of realistic knowledge during planning.
There are two outputs from the activity duration estimating process group:
Top of Page