# How to determine durations correctly in MS Project?

## How to determine durations correctly in MS Project?

Do you determine durations correctly in MS Project? Why are we not finishing all our projects on time? Do we actually calculate correctly from beginning on when setting up the tasks, durations and their estimates?

Get our ‘best practices template’ now!

Save yourself and your whole project team time and money by giving your tasks the right duration. So how do you do that? Let’s start at the beginning:

From the moment you have all your tasks structured in your schedule, it is time to give these task the right duration. At this point, consider the number of working days (exclude weekends) between start and finish of the task. This means that if you work two hours per day on a task for five consecutive days starting Wednesday, the duration will be 5 days.

### Adding durations to tasks in MS Project can be done in two ways:

1. By typing the number of days into the Duration column
2. Have Duration be calculated by MS Project by assigning resources and their availability to the task and estimating the hours of ‘work’.
3. This is explained in our e-course ‘Resource Management‘, as many factors will then determine your duration.
You will see that sometimes the duration has a question mark. This appears when the duration column has not been edited. The idea is that as long as this value is not put in by the user, it is an estimated value.

Durations of tasks and thus projects are determined by the amount of effort they take and the availability of resources that spreads this effort over the duration of the task. This means that whether you put in the duration in your schedule directly or have duration be calculated by MS Project, duration is ALWAYS effected by the total estimated amount of effort and how the resource is assigned to the task. After agreement on resources and on budget, and this budget is correctly reflected in the availability of resources in your schedule, the process of effort (‘work’ in MS Project) estimating becomes the most important factor to get reliable lead times.

One of the prerequisites for producing the best estimate is to have spent sufficient time analyzing and understanding the requirements and the proposed solution.

Managing project durations is often a challenge due to (the great number of) uncertainties. Planning without regard for these uncertainties would induce risk of not meeting the agreed plan date. Plan dates depend on information we know – the agreed triple constraint (scope, time and resources) as well as things we do not know – actual scope, actual effort, actual resource availability and risks (based on agreed changes). To include contingency time for 100% of all likely and unlikely events is not wanted. Therefore the level of risk you accept to take largely determines the estimates of the effort. At the next chapter we discuss how you should factor in the risk in your estimate.

### Level of detail

The level of detail you choose determines the accuracy, and thus the quality of the estimates. The notion that detail is needed to make more accurate estimations is NOT correct. When planning tasks of an hour, the likelihood that you will exceed by 100% is much greater than planning tasks of for instance 40 hours. For planning with as accurate as possible estimations 16 – 40 hours per task is ideal.

On the other hand, the validity of the estimate largely depends on how much time in advance the estimate was made, regarding the simple fact that uncertainty increases when you look further into the future.

As a guideline: plan tasks of about 16-40 hours (can be spread over a longer duration) for the next 3 to 6 months. Every task after this period could have less detail as many things could still be unknown.

### Estimating effort (work)

There are many factors that determine the estimates:

• Experience
• Avoiding risk of failure
• Wishful thinking
• Pressure from stakeholders to shorten the plan
• Pressure to come up with a plan fast (not enough time to perform estimation)
• Experience of colleagues
• The complexity of the tasks
• Are the deliverables clear?
• Available knowledge
• Available FTE and resources
• Etc.

### The best practice for coming to estimates is:

• Select a group of 6-8 of the most knowledgeable people with regard to the topic for the session, for instance engineers, stakeholders, certain team members and if known the resource that will be doing this assignment. These experts do not need to be from the own project team. Since many times the actual resource is not yet known, the experts need to consider a generic resource (a role) – how many hours would the average person in this role (with this competence) need. Later the estimate can be reevaluated by the actual person doing the task.
• Prepare the session as good as possible by knowing the deliverables and their requirements, having the necessary documents available, and have specifications and must-have results aligned with the sponsor.
• In the session: Align the tasks (every person has the same understanding of what needs to be done) and have every person give their best estimate for the tasks.
• Plot the 6-8 estimates in a graph like below. Less estimates means less combined expertise and will decrease the accuracy of the estimate.

• Discuss the estimations all participators made and learn from each other which assumptions were done making the estimates. Start with the persons giving the highest and lowest values (1 and 5 in above example). Make sure each participant knows why choices are made.
• Reach consensus on:
1. Best case (1,3,6)
2. Most likely case (6,4,8)
3. Worst case (2,7,5)
• Repeat these steps for all tasks that are part of this estimation session.
• The estimate is calculated like this:

Best case                      :           16 hours          x          1          =            16

Most likely case         :           65 hours          x          4          =          260

Worst case                   :           120 hours        x          1          =          120 +

Total                                                                                               =          396 hours

Divide the total by 6.
396 hours / 6 = 66 hours

66 hours would be a realistic and reliable estimation for this task to use in the column ‘work’ when you create your schedule.

Note: Document the estimations and assumptions which are discussed with the responsible resources to create a reliable estimation database.