Before a schedule can be produced, we need to know what resources and time will be required to complete each activity. These need to be estimated.
Once it has been decided what products need to be created and in what sequence, we are in a good position to start considering the resource inputs required. We should recognise, however, that this is an iterative process and the results will depend on what tasks need to be completed and who will carry them out.
When estimating time, we should be clear as to what we mean by ‘time to complete a task’. We might consider time in two ways:
- Actual time - this is the time we would book against a project - the actual time spent completing tasks to meet the project objectives. Project Leaders need this information for job-costing.
- Elapsed time: this is the time a task would take from start to finish and may include time working on other non-related activities. Project Leaders need this information for project scheduling.
If, as Project Leader, you ask a team member how long it would take to prepare a job description for a new member of the team, you might be told two hours, but because of other work commitments, it won't be completed for three days. The two hours is the 'actual time', and gives an indication of the cost of that piece of work, while the 'elapsed time' is three days, which tells you that any dependent tasks, such as advertising the job, cannot start before that time.
Factors that need to be taken into account when estimating actual and elapsed time, therefore, are:
- Other work commitments
- Experience of the person/function/department
- External events or influences, e.g. holiday periods, supply dates
- Additional time for quality control activities or obtaining approval
- The risk of things going wrong.
'Accurate' estimate is a contradiction in terms. An estimate cannot be totally accurate as it is a guess. However, there are a number of techniques that can help produce estimates on which we can base our planning with a degree of confidence:
- Ask the people who will actually do the task (if you are using subcontractors, they may be contractually bound by this estimate)
- Get the opinion of an expert
- Find a similar task in a completed project
- If it is a task that will be repeated several time, conduct a test activity
- Make an educated guess (not an arbitrary one).
The Project Management Institute’s ‘A Guide to the Project Management Book of Knowledge - Third Edition’ (PMI, 2004) identifies a number of tools and techniques for estimating time and resources. They include:
Analogous Estimating - This means using the actual duration/cost of a previous similar activity as the basis for estimating the duration/cost of a future activity. This is frequently used when there is a limited amount of information available about the detail of the project. The technique relies on the availability of historical data and expert judgement. Clearly this is most reliable for new projects that are very similar to those used as a comparison.
Bottom-Up Estimating - When the duration or cost of an activity cannot be estimated with any degree of confidence, the work within the activity is decomposed into more detail and duration/cost estimated for each. These estimates are aggregated to give an overall estimate.
The size and complexity of the overall project obviously has an effect on the reliability of this method as the interdependency of tasks can influence the way resources are used.
Parametric Estimating - This method involves estimating the time or cost of doing a task once and then extrapolating to calculate the total time or cost of doing the same task repeatedly. Overall duration is calculated by dividing the total time by the number of resources applied to the task.
Three Point Estimates - In estimating the time required for a task, we could assume that it would fall somewhere between flawless execution and major disaster. The best approach is to establish a compromise between the two. Three points estimates require estimation of three scenarios:
- Optimistic estimate: no problems are encountered
- Realistic estimate: a few problems crop up and introduce delays
- Pessimistic estimate: major problems occur that seriously jeopardise the task but the project will still go ahead.
As a Project Leader you might be tempted always to choose the pessimistic estimate for each task. Or indeed, you might simply decide to take the optimistic estimate and add an arbitrary 20% to cover contingencies. At least this way you are unlikely to exceed budget or overall timescale.
However, there are problems with this approach. If you have been given a deadline or a budget for the project, you might not be able to fit all the tasks within the constraints. The only option you have open then is to reduce the scope of the project. This means the project will not achieve everything it set out to achieve. Alternatively, if the project is open-ended in terms of timescales or budget, you might be tying-up resources needed for other projects. Remember, 'work expands to fill the time available'.
Obviously, if you always choose the optimistic estimate, you are just asking for trouble.
Three points estimating produces a form of 'weighted average'.
Where a, b and c are the weighting factors. Typically these are set to a=1, b=4, c=1. However, for projects where there is high degree of uncertainty, it may be appropriate to adjust the weighting factors in favour of the pessimistic, e.g. a=2, b=1, c=4.