Using data from similar projects to estimate your project activity durations is called

All that you need to do now is to figure out the effort required for each of those activities, right?

No, I’m afraid.

This point can be a bit confusing. Read on.

Estimating Activity Durations process is about figuring out the effort required to complete each activity, AND the amount of resource required for the activity, AND figuring out how long does it take to complete the activity by combining these two.

Let us look at an example –

As we saw earlier John is managing a project of building a house for Josh. One of the tasks he has for the laying foundation is ‘Pour concrete footings’. John knows that for this activity he needs some amount of concrete, and a few people to pour the concrete into the trench.

In the project management activity of Estimating duration of Activities he will need to figure out the quantity of concrete and the number of people he needs for this task. Let us say that would be 800 cubic-ft of concrete and 6 people. He calculates that it takes 4 days for 6 people to pour and set 800 cubic-ft of concrete into the trench. Hence the duration for this activity would be 4 calendar days.

Duration is in calendar days (or hours, weeks – the unit of work chosen for the activity), and not raw the effort like number of person days. In short, Duration is not raw Effort.

Concepts that influence activity duration estimation

While estimating activity duration seems straight forward, there are few concepts that can impact the estimates even when you go by the book to estimate them. Let us look at a couple of them –

Warren Buffet once said, “You can’t produce a baby in one month by getting nine women pregnant.”

Choice of technology may have an influence on the duration of an activity. In manufacturing, for instance, usage of robotics or CNC machines can have substantial difference in the time it takes to complete certain tasks. In case of software product development, for instance, choice of framework or language can make a world of difference on activity duration.

Finally, motivation forms an important part on adhering to schedule. Lack of motivation does not get the work done even when the person has all the required skills, facilities, and support to get the work done.

Now, what goes into estimating duration of an activity?

You have seen most of these inputs already in previous project management activities – schedule management plan, list of activities, attributes of activities, and resource requirements for activities. The last one is required to know the level of skill needed for each activity because a lower skill-level person assigned to an activity will have overhead such as communication and oversight, and potentially have reduced productivity.

Recall from the project management activity to estimate resource needs of an activity that resource calendars contain information such as availability, skill level, type, and quantity of people and equipment in the organization. Now we have used this while preparing activity resource requirements anyway, why consider it again now for estimating durations?

Very valid question.

If you work in software industry like I do, you would know that it is extremely difficult to measure productivity of people. Two engineers with same number of years of experience are likely to take different amount of time to write the same program, and they write it with varied degree of quality of code. In addition, a junior engineer has far higher chances of writing software that requires checks and code-reviews.

Complexity of module will take this effort only exponentially up. This kind of information is useful while calculating activity duration. That is why we need resource calendar while estimating durations.

Any contractual terms and restrictions, and any documented constraints and assumptions are considered from project scope statement.

Risk register may contain any potential risks to be considered while estimating activity duration, and need to be looked at.

Resource breakdown structure is a hierarchical structure (as in work breakdown structure or WBS) of resources by resource category and type.

This is how it’s done!

Expert judgment

This is effective in this process when you involve team members. Consider the software development case – who is a better judge of the effort and duration than the person who is likely to do the work? In spite of involving team members you may need to involve other experts in the company, or even from outside for critical or complex tasks.

Involving the team will give it a head start, gets you more accurate estimates and most importantly – bring up any gaps, dependencies and issues lying dormant among activities.

This probably is the last gate to catch any of these risky gaps, because schedule will be prepared after this and release timelines will be calculated. Once schedule is baselined, any changes will trigger change request process. It is way much better to keep an eye for challenges at this stage and include durations accordingly.

Analogous estimating

This is pretty straight forward. You look at similar activities with similar resource category and types from similar type of projects executed earlier and base your estimates on this information. This method is most useful when you don’t have enough information in the current project and you are doing progressive elaboration.

This technique takes lesser time but at the cost of being less accurate. Best fit for this is the case when current project and the team is similar to the previous project and its team. Hence, it is better to use this technique in combination with other tools and techniques.

Parametric estimating

This is an activity durations estimation technique where statistical techniques are used to calculate cost or duration values for activities based on data from similar earlier projects. You put in numbers, or parameters, into a spreadsheet or software that uses data from other similar past projects to calculate estimates.

For instance, per historical data it takes a worker one hour to dig 50 square ft. of land using a certain earth digging equipment. Using this data John calculates that for his house project of 800 square ft. it takes about 16 hours (or two 8-hr days) to dig the earth.

Parametric estimating method gives more accurate results than analogous estimating technique provides.

Three-point estimating is a technique that takes uncertainty factor into consideration.

Consider this example.

You ask a software developer how long a program will take to write, most likely you will get a number that is pretty optimistic. To get realistic estimate you would ask him to think of worst-case scenario (pessimistic estimate, tP), best-case scenario (optimistic estimate, tO), most-likely scenario (tM), and estimate durations for each of them. Then you run the following formula to calculate a number that is closer to reality.

Expected duration = [Optimistic duration + (4 x Most likely duration) + Pessimistic duration] / 6

tE = [tO + 4tM + tP] / 6

Let us go back to John’s house construction project. John figures that the task ‘Lay drainage tile system’ has an uncertainty due to the soil condition which may not hold the tiles in place. He talks to the team and asks for best-case, worst-case and most-likely estimates. Members put their heads together and come up with 12, 24 and 16 days respectively.

John works out the expected duration to be = [12 + (4 x 16) + 24] / 6 = 16.67 days, rounded off to 17 days.

Now above formula assumes Beta distribution [from Project Evaluation and Review Technique (PERT)] to be the distribution of values within the range of three estimations.

There is another formula, based on Triangular distribution.

Expected duration = [Optimistic duration + (Most likely duration) + Pessimistic duration] / 3

tE = [tO + tM + tP] / 3

Group decision-making techniques are all those team-based discussions that help people arrive at a rational decision. Examples are brainstorming, Delphi or nominal group techniques. More you involve people in making decisions more will be their commitment to project success.

We saw this technique in Collecting Project Requirements lesson.

Reserve analysis

This is a form of Data analysis, and is quite simply the buffer you would include to take care of uncertainty. There are two types of reserves –

• Contingency reserve – this is for those uncertainties that are identified in risk register. Hence contingency reserve is for “known-unknowns”. This is usually given for specific tasks if they are complex, or added as a percentage of overall estimates. This is part of schedule baseline.

• Management reserve – this is not included in the schedule baseline. Management reserve is calculated in terms of percentage of project schedule duration and available only for management’s control. This is for discovered work that is within project scope. In other words, management reserve is for “unknown-unknowns”. Because this is not part of schedule baseline, the schedule might need to be baselined again if and when management reserve is utilized.

Click herefor more details about ‘knowns and unknowns’ and an interesting reference to this by United States Secretary of Defense Donald Rumsfeld..

Both these reserves must be documented clearly in schedule documentation.

Alternatives analysis – there are multiple routes to the same destination, and this analysis technique tries to explore alternate, better solutions to do an activity that reduces duration.

Make-or-buy-or-rent-or-lease analysis, manual versus automated approaches, schedule compression techniques are some of the ways to do alternatives analysis.

Exam pointer: PMBOK® guide uses a generic word for unit of measure for duration – work period.

What really do you get in the end?

Activity duration estimates

Estimates are approximations, which means it is hard to give an exact number with 100% certainty. However, as a project manager your intent is to give a reasonably confident estimate, with a certain range.

Consider John’s project to build a house – he has a work package called ‘lay foundation’, which he has broken down into few activities such as dig ground, pour concrete footings,lay sub-slab system. When he estimates them and rolls up the estimates, he gets about 19 calendar days. Based on reserve analysis and expert judgment he arrives at an estimate of 19 + or – 2 days as estimate. This means he has a range of 17 to 21 days to complete this activity.

Activity duration estimate is the work period for a task expressed in numbers or in some case as a range. However, there is neither any explicit buffer included, nor leads and lags.

As you can see, reserves can be given at activity level or work package level. Some project managers keep a contingency reserve at milestone levels to derisk the published milestone releases. This however depends on the risk appetite of the organization and customer’s urgency around milestone deliveries.

In Critical chain method a ‘feeding buffer’ is given at points where a chain of tasks join critical chain’s path, so that a reasonable delay in the chain does not impact critical chain.

Each time you do something with activities you get to know more about them. If an activity is as simple as ‘hiring an excavator’ it may take few hours. However, an activity like ‘laying drainage-tile system’ may need the team to think through all that can go wrong, in order to estimate it well. And each time you think about it, you may find out something more about the task. You may even discover some implicit assumptions.

Where will you document this information?

In activity attributes. There could be other documents as well that may be updated during this process. Hence the other output of this process is project document updates.

There you go!

You have everything needed to produce something that most of the stakeholders love to understand. The Project Schedule. That is the content of next lesson. Take a break, have a cuppa coffee, and jump right into the next lesson.

It only gets better from here.