Program Evolution and review technique (PERT) is a estimation technique which is used to find out the project duration or the project cost.
Before going to the details of program evolution and review technique (PERT), let us understand what is this estimation thing all about.
When the project manager identifies the project tasks he or she first needs to decompose the tasks into the smaller activities. Then the project manager needs to find out how much time requires for each of the activity to come up with the project estimate.
But the question is how does the project manager knows how much time it takes for each activity in the project?
Single point estimation
The project manager needs to work with project team to find out if there is any historical data on the basis of which we can estimate the duration for each activity. This would definitely provide a basis for the estimate.
For example if an similar (or same) activity in a historical project took 2 weeks of time to complete the activity, this estimate can be used in the current project activity as well to calculate the project duration.
This is called single point estimation. However in all cases, we may not be able to find the activity duration from historical projects.
For example the project team has never done this kind of activity earlier. Or the project team is not sure at all how much time “exactly” the activity takes.
In such cases, the three point estimate will give better results.
Three point estimate
Three point estimate is of help in cases where the project team does not have any historical basis for the activity estimates.
In these cases the project manager sit with project team to understand three different estimates for each activity. They are
- Optimistic (O) estimate
- Most Likely (M) estimate
- Pessimistic (P) estimate
Let us look at each of them with the below example.
Three point Estimate Example:
For example if the project says that this particular activity is going to take usually (normally) 10 days of time considering the nature of the activity. In the best case the activity is going to take 6 days for completion, if everything goes smoothly. In the worst case it would take 15 days to complete the activity.
Optimistic Estimate
When the project team is not fully aware of how much time exactly the activity takes, then they try to see what is the range that they can be suggest to the project manager.
Optimistic estimate is always a best case estimate in that range.
In the above example 6 days is the optimistic estimate.
Most Likely estimate
This is the most likely estimate for a specific activity. Meaning that the project team thinks that in most probable cases, this activity should complete within this estimate.
In the above example 10 days is the most likely estimate.
Pessimistic Estimate
This the worst case estimate. Meaning that considering any unforeseen problems, even in worst case the project team can complete this specific activity within this pessimistic estimated duration.
In the above example the pessimistic estimate is 15 days.
Types of three point estimates
Now with the help of all these three estimates optimistic (O), most likely (M) and pessimistic (P), we can calculate the three point estimate, which is proved to be more accurate than the single point estimate.
There are 2 popular ways of calculating the three point estimates. They are
- Simple Average estimate or triangular distribution
- Program Evolution and Review Technique (PERT) or beta distribution
Simple Average estimate or triangular distribution
When you plot this three estimates on a graph, then it looks like a tringle and hence the name triangular distribution.
Simple average estimate or triangle distribution is calculate by taking a simple average of the three values optimistic, most likely and pessimistic estimates.
Triangular distribution Estimation = (Optimistic Estimate + Most Likely Estimate + Pessimistic Estimate)/3
This will definitely provide a better result than the single point estimate. However the only problem is that triangular distribution considers the same weightage for all the 3 estimates. Meaning that for optimistic, pessimistic and most likely estimates are taken by simple average, without any weightage.
Program Evolution and Review Technique (PERT) or BETA distribution
On the other hand program evolution and review technique (PERT) uses more weightage to the most likely estimate compared to the other two estimates pessimistic and optimistic.
This is because most like estimates has more probability than the other two. Hence this is even more accurate than the triangular distribution.
We can calculate PERT as follows:
PERT Estimation = (Optimistic Estimate +4 x( Most Likely Estimate )+ Pessimistic Estimate)/6
In the above formula most likely estimate has given more weightage (4) than the other two (pessimistic and optimistic).
Three Point Estimate Example
Let us take a simple example of how to calculate the 3 point estimates.
An activity pertaining to a construction project takes 5 weeks in best case, 7 weeks is the most likely case and 12 weeks is what it takes for the worst case.
So if we calculate three point estimates for this example
Triangular distribution estimate = (5+7+12)/3 =8 weeks
On the other hand the Program evolution and review technique or PERT estimate = (5+(4*7)+12)/6 = 45/6 = 7.5 weeks
PERT and CPM (Critical Path Method)
PERT and CPM are two different type of methods. Program evolution and review technique (PERT) is used when there are uncertain or unknown activities in the project. On the other hand critical path method considers well defined project activities.
PERT is a probabilistic model whereas CPM is a deterministic model.
Generally PERT is more suitable for R&D type of projects, where the activities are done by the project team for the first time. On the other hand CPM (Critical path method) is used in non-R&D projects, where the activities are well established.
Finally PERT greatly helps to find out activity duration of a project, while CPM helps in understanding the critical path in the project.
Conclusion
To summarize, we have seen what is a single point estimate and is when is it relevant to use.
What are three point estimate is and different types of three point estimates.
Importantly we have seen how to calculate program evolution and review technique (PERT) and also the triangular distribution.
Finally we have ended the topic with a small comparison between PERT and CPM.