Estimating task durations is one of the fundamental parts of project management. It involves the estimation of the amount of time required by a certain project activity given the available resources. This happen directly after cost estimating. In the Project Management Body of Knowledge (PMBOK), the main output of the Estimate Activity Duration process are Activity Duration Estimates.
This is a critical piece of the puzzle. If you underestimate the duration you can cause yourself substantial headache later on when the project goes behind schedule.
The units of the duration can be whatever works best, but usually hours or days will suffice. For small projects less than about 2 weeks in duration, tracking in hours works pretty good. Larger projects will not want to be bogged down by such small units and will need to be tracked in days.
Activity Duration Estimates
The main output of the Estimate Activity Durations process is a duration for each task. When manually developing a project schedule, each activity requires a duration, like this:
|120||Build Forms||2 days|
|130||Place Rebar||6 days|
|210||Pour Concrete||1 day|
|310||Setting & Curing||5 days|
|320||Strip Forms||2 days|
This is called Bottom Up Estimating, whereby each task is given an estimate and the total is rolled up into the overall project estimate.
The units don’t matter – They could be hours, days, weeks or even production units like volume of concrete poured. But time based units are the most intuitive and best for future planning. My advice is to use days if possible, because:
- Project deadlines are normally expressed as a certain day of the month. If all your planning is in hours, weeks, or some other unit, you will need to be converting to days constantly to ensure you meet the deadline.
- The gantt (bar) chart in the next step will use days as its primary unit.
The duration estimates can also be a range, for example 4-6 days, or a probability, for example 4 days @ 90%. But I would not recommend this type of scheduling unless you are a professional and understand the use of earned value analysis and project control with variable activity durations.
Some people insist that the people doing the work are the best estimaters. I generally support this view, but this is often not practical and project team members do not want to be bogged down with endless duration estimates. The project manager can, and should, learn how long the key activities take (as well as how much they cost).
The quality of the estimates always goes up with time, eventually morphing from an estimate into an actual duration. At the outset, the project manager might need to insert contingencies to account for this uncertainty, and the contingencies decrease as the task is performed.
How to Estimate Durations
Estimaters have six main tools in their arsenal when estimating activity durations. To estimate an activity duration, proceed through this list and perform each one, skipping the ones that do not apply to your project:
- Analogous Estimating
- Parametric Estimating
- Expert Judgment
- Three-point Estimating
- Group Decision Making
- Reserve Analysis
As the name suggests, this is when you make an analogy to the same, or similar, task that has been performed before. This is the best source of information because actual work completed, even if it requies adjustments, is extremely reliable. For example, if your project is for building a driveway and you’ve done it before, you clearly have a head start. Often, however, you have to make some adjustments.
- The task is bigger or smaller than the last one.
Scale the duration up or down by the appropriate amount. For example, this driveway is 20% larger than the last one, so maybe the overall duration of the concrete pouring task will be 1.2x the last one.
- The task needs specialty products or services that the last one didn’t.
In this case, add in the appropriate products and services and proceed as usual.
- The task is for a slightly different end product.
Determine the difference in materials and/or workmanship and figure out how many hours or days more or less it would take.
The productivity of the work is an important consideration. For knowledge workers such as engineers, adding more sometimes slows the production and decision making process down instead of speeding it up. Likewise, a very experienced laborer has higher productivity than a less experienced one.
A second common method for estimating task durations requires that the work be drilled down into a unit cost, such as a time per square foot of concrete pouring. Here are some examples:
- Pouring time per cubic yard of concrete.
- Total time (forming, rebar, pouring, and setting) per cubic yard of concrete.
- Installation time of carpet per square foot.
- Ultrasonic inspection time per square inch of rotor blade.
In the engineering industry almost everything is done this way, from the engineering time down to the construction materials.
Sometimes a parametric value includes fixed items, such as in #2 above, where the total time to pour a concrete slab includes the rebar, forming, and setting. Obviously some of these variables do not correspond linearly to the overall duration, but the estimate is considered “close enough” that it will work.
Professional estimators will tell you that in spite of all the techniques in the book, a technical expert will be the best resource you could ever have. They could be the only resource you need for estimating. Technical experts are notoriously busy because their expertise gets used on many projects. But if you have access to one, whether inside the company or out, you should find a way to use them.
But what if a technical expert contradicts the other methods? For example, let’s say you’re installing carpet and your carpet installer tells you a certain project will take 4 days, but past experience (analagous) is telling your it will take 10 days. How valuable is the expert’s opinion then? It is difficult to give good advice that applies to all situations. You must simply weigh the two methods against each other, considering the underlying data of each method. For example, maybe this carpet installer frequently underestimates duration, or the past experience estimates are skewed by a certain piece of data.
Project management itself does not exist independent of the technical expertise. That is, the project manager or management team must be at least passingly familiar with the technical aspects of the work, or the project has little chance of success.
In this method, the estimator determines three numbers:
- Most Likely
The second, “Most Likely” is the average of what the task duration would be if performed many times. The other two simply represent the 90% confidence interval for the task. You can ask functional managers or technical experts how long a task would take once every 10 times you perform it. By the way, the methodology for this varies. Many textbooks use 99% confidence, or some other system. Suffice it to say it should be an optimistic vs. pessimistic estimate.
Using the beta distribution, the expected duration is then:
te = (a + 4m + b) / 6
- te = Expected duration
- a = Optimistic
- m = Most likely
- b = Pessimistic
For example, let’s say the expected duration of the task Place Rebar is 6 days (m = 6), and you figure there is significant potential for delays. The Optimistic estimate is 5 days (a = 5), but the Pessimistic estimate is 13 days (b = 13). The beta distribution yields 7 days as the task duration you should use.
The standard deviation of the task is simply:
σte = (b – a) / 6
If you add up all of the durations (∑m) and standard deviations (∑σ) for each task, you will have the total project duration as well as it’s standard deviation. You can draw the following conclusions from this:
- The confidence level in one standard deviation, i.e. M ± σ is about 68%.
- The confidence level in two standard deviations, i.e. M ± σ is about 95%.
- The confidence level in two standard deviations, i.e. M ± σ is about 99.7%.
In the example above, the standard deviation is (13 – 5) / 6 = 1.33. Therefore, the task duration is:
- 7.33 days at a 68% confidence level
- 8.66 days at a 95% confidence level
- 10 days at a 99.7% confidence level
Of course, this exercise in statistics is dependent on good estimates of the original input data, which is the Optimistic/Most Likely/Pessimistic estimates.
Group Decision Making
This technique attempts to utilize the collective wisdom of groups, thereby eliminating individual biases. There are multiple ways to do this.
- Consensus Decision Making. This technique requires a majority of the group to determine the decision. You could present three options for a task, say “Concrete pouring” could be either 4, 6, or 8 hours. The majority opinion determines the duration estimate.
- Weighted Vote. Alternatively, group members could vote with a score of 1, 2, and 3, and the results weighted to determine the highest score.
- Delphi Method. In this method each group member is asked to make an estimate independently without knowledge of the others. Then the results are compiled and shown to the group without identifying whose estimate belongs to who. The group tends to converge very quickly to one single result.
Theoretically, each task contains certain risks which, together, are agglomerated into an overall task risk amount. For example, the concrete truck might not arrive on time and delay the concrete mix by 1 day. If the chance of delaying the concrete pouring by a day (plus all the other risks) is great enough, a task duration of 2 days or more days should be used. The second day is called contingency reserves and is planned into the schedule. Overall, the project can be assigned a management reserve which is available account for unidentified, unplanned risks. When the concrete truck arrives 4 days late, the schedule is adjusted and additional costs incurred are taken out of the management reserve.
You need to make sure the entire possible span of the work activity is included in the estimate. For example, it might take 24 hours to place rebar, but maybe you want to spread it over 12 days instead of 6, to allow for some contingencies.