One of the first mistakes that new project managers make when carrying out estimate activity durations is to assume that they must estimate the activity duration of each task. This is called ‘guessing’!
Put simply, every activity has three attributes; its duration, the work effort and the resources. The secret is to estimate the work effort first, then the resources, and finally be able to estimate activity durations. Estimating and activities duration straight off the bat is a bit like determining the price of meal before deciding what you want to eat and where you want to eat it!
The Estimate activity durations process within PMBOK is to analyze each activity and estimate how long it will take, but as you can now see, it will depend on the amount of work required within the activity, how many resources will be assigned to this activity, who those resources are, and when they are available.
You will notice that the sequence within the time management process group is to first define the activities, then estimate the activity resources and as we are doing here, then go on to estimate the activity durations.
Activity list. This is the main input and ensures that every identified activity is included so that its duration can be estimated
Activity attributes. These always accompany the activity list and provide extra information about each activity such as the number of resources needed for each activity.
Activity resource requirements. This was a main output from the estimate activity resources process and includes the kind of resources and their number along with the reasons that led to those choices
Resource calendars. It is worth mentioning that resources can be human or non-human such as facilities, materials, tools and equipment. Human resources have standard working hours and hence are constrained by their availability. As an example, if a task requires 50 hours of work, then its estimate activity durations will be based on the number of people to carry out those 50 hours of work, factored by their calendar availability.
Project scope statement. This is defined and created within the define scope process and includes any assumptions and constraints which will have a bearing on estimating the activity durations. As an example, the scope statement may identify them as a particular activity needing to be subcontracted due to the specialist nature of the skills required, and this would assist in obtaining an accurate using the estimate activity durations process.
Organisational process assets. These may take several forms, but the main one will consist of historical information for previous similar Estimate activity durations. These assets may also consist of information about the work culture within the organization and typical working patterns such as overtime or shift working.
Enterprise environmental factors. The nature and environment of the project will have a large impact on the time it takes to carry out certain tasks. If this task were taking place on an oil rig, then stringent safety standards and the difficulty of carrying out such an activity in the middle of the ocean and under hazardous conditions, would have a bearing on Estimate activity durations.
There are two main outputs from the estimate activity durations process:
Estimate activity durations – Activity duration estimates. This of course is the main output from this process. Do not assume that all activities will have one estimate as there are many techniques and tools helping to give a range of estimates for each task. I will develop this idea further when discussing the tools below.
Estimate activity durations – The project document updates. As a result of carrying out an estimate it is important to capture any assumptions that they have been made for example the skills knowledge and experience required. It is very likely therefore that the activity attributes will need to be updated.
There are four main tools used in Estimate activity durations:
Estimate activity durations – Analogous estimating.
This uses estimates from any previous similar projects as a basis for estimating activities within this one. These previous estimates are arrived at using expert judgement so that the amount of similarity from the previous project to the current activity can be factored in.
Suppose you and your family are about to go on a camping holiday for the first time. You practice directing your tent in your backyard and establish that it takes you 30 minutes. Knowing that you will direct your tent halfway up the mountain where the temperature will be lower and strong winds will be present, you factor this in and estimate it will take you 45 minutes.
Analogous estimating is also known as top-down estimating.
Estimate activity durations – Expert judgement.
Using individuals who have performed similar activities before it is a powerful way to obtain realistic and achievable estimates. The knowledge, skills and experience they have gained will improve the accuracy of such estimates. Indeed, without any knowledge of the activity, then any estimate would at best, be a simple guess.
Expert judgement is best implemented in parallel with organisational process assets and consideration of the activity environmental factors.
Estimate activity durations – Parametric estimating.
Put simply, this is linear extrapolation, and hence best applied for activities that can be easily scaled. A prerequisite of using parametric estimating must therefore be that experience of similar previous activities is available.
Suppose that it takes an engineer 2 hours to configure a notebook PC. If 20 personal computers need to be configured, then it will take 40 hours of work. If it is realistic to assume that the same engineer can configure for personal computers every day, then it will take five days to configure all 20 computers. Put another way, five engineers with similar knowledge skills and experience, could configure all 20 personal computers in one day.
Estimate activity durations – Reserve analysis.
This is an approach used by risk management and may also be called contingency. This must not be confused with the term ‘padding’ as this would infer adding ‘fat’ to any estimate.
The approach here should be to consider any risks that may occur when carrying out a particular activity and the impact such risks would have all the activity completion. Examples of such risks might be rework, errors, technical difficulty, resource availability, etc. Based on such risk analysis, a contingent amount would be added to the activity. For example, if reworked was a risk and likely outcome, and it was determined than an extra 5 hours work would result, then either more resources would be added to the activity, or its duration increased by an extra day.
The risk management is an ongoing activity throughout any project, and it is important therefore but such reserve analysis time additions should be reviewed throughout the project and revised as appropriate as more information is gathered throughout the life of the project.
Three point Estimates.
This technique is also known as PERT (project evaluation and review technique) estimates.
If only one estimate were to be used for a task, then there is a high probability that the estimate will turn out to be slightly higher or lower than reality. But suppose when seeking an estimate, that you asked for three estimates instead of just one.
The approach here, is to ask for a pessimistic or worst case estimate, a realistic almost likely case estimate, and an optimistic or best case estimate. You may wish to use work hours in carrying out this technique and then convert them to duration. Alternatively, you may use this technique directly with duration’s such as days:
The three estimates stated above are replied in a formula that produces a weighted average of the three and results in a single more realistic estimate for the task duration.
Another formula to remember is that of standard deviation for each estimate is that standard deviation = (pessimistic – optimistic) / 6
Using a simple example, let’s assume the following estimates were obtained:
This would result in the following 3-point estimates:
Applying the standard deviation to each (which is merely a factor of the ‘spread’ of the estimates) will give a rough guide to the confidence you should have in each estimate (a lower value being better).
David spent 25 years as a senior project manager for US multinationals and now develops a wide range of project-related downloadable video training products under the Primer brand. In addition, David runs training seminars across the world, and is a prolific writer on the many topics of project management. He currently lives in Spain with his wife Jude.