Locked lesson.

Upgrade

About this lesson

Project plans are built with an accumulation of estimates, each of which has a level of uncertainty associated with it.  The level of uncertainty is a major contributor to the accuracy of the plan and the amount of project risk.

Quick reference

Estimating Uncertainty

Project plans are built with an accumulation of estimates, each of which has a level of uncertainty associated with it.  The level of uncertainty is a major contributor to the accuracy of the plan and the amount of project risk.

When to use

When estimating any aspect of project tasks, determine what type of uncertainty exists and then apply the appropriate estimating strategy. 

When preparing or revising estimates to complete the project, determine what type of uncertainty exists and then apply the appropriate strategy.

The estimating strategy can be applied to cost estimates, time duration estimates, or level of work estimates.

Instructions

  1. For each task or task group, determine which category best describes the task effort on this project:
    1. Basic – A task that is a core competency of the organization, little uncertainty.
    2. Dependent – A task that is well understood but the estimate is based upon some factor that is not yet certain.
    3. Uncertain – A task that is not well understood, there is a high degree of uncertainty.
    4. Unknown – Aspects of the project for which the team has no information, there is total uncertainty.
  2. Depending upon the nature of the uncertainty, create the estimate.
    1. Basic – Use standard project estimating techniques.
    2. Dependent – Create an estimate of the task for each of the states of the dependent factor.  Select the state that is most likely, but show the factor as a risk and monitor it as such.  If the factor changes, the estimate needs to change.
    3. Uncertain – Estimate the best-case, worst-case, and most likely case for the task.  Normally, use the most likely in the project plan, but build project reserves based upon the best case and worst case.  Typically create a reserve that would cover about 75% to 80% of the worst case.  This task should be tracked as a risk regardless of the estimate used.
    4. Unknown – The estimate is a guess and the area should be tracked as a high-risk aspect of the project.

Hints & tips

  • Ignoring uncertainty does not make it go away, so acknowledge it and manage it.
  • Uncertainty factors can often be managed so as to bias the actual time and money spent on a task to be close to the estimate.
  • The majority of the tasks, even on high-risk projects, will be basic tasks.  Have your team use standard estimating techniques on these and focus your project management attention on the other tasks.
  • Uncertain tasks can often be decomposed into smaller tasks – many of which will likely be Basic tasks.
  • When working on project activities with unknown tasks, do frequent pulsing checks with the project team to learn factors that can help you create an estimate.
  • Definition of Estimate: “A quantitative assessment of the likely amount or outcome of a variable, such as project costs, resources, effort, or durations.”  PPMBOK®Guide

This definition is taken from the Glossary of the Project Management Institute, A Guide to the Project Management Body of Knowledge, (PMBOK® Guide) – Sixth Edition, Project Management Institute, Inc., 2017.

Login to download
  • 00:03 Hi, I'm Ray Sheen.
  • 00:05 I'd like to talk about estimating on projects, and in particular,
  • 00:08 how to create estimates when there's a great deal of uncertainty
  • 00:12 concerning some of the project tasks.
  • 00:15 The Project Management Body of Knowledge, the PMBOK guide,
  • 00:19 defines an estimate as a quantitative assessment of the likely amount or
  • 00:23 outcome of a variable, such as a project cost, resource, effort or duration.
  • 00:29 So what does that mean?
  • 00:30 You need to have a number but
  • 00:32 with a project estimate that number is likely to have uncertainty.
  • 00:36 Often you may qualify the estimate with terms such as a preliminary,
  • 00:41 feasibility, order of magnitude estimate to convey this uncertainty.
  • 00:46 And sometime, you qualify the estimate by putting tolerances on it,
  • 00:51 such as plus or minus 10%, or confidence levels of 70% or 80%.
  • 00:56 These are usually judgement calls and not precise calculations.
  • 01:00 The uncertainty normally depends upon the type of activity associated with the task.
  • 01:05 So let's take a look at the task uncertainty as we consider our estimating
  • 01:09 approach for each project task.
  • 01:12 One type of task is a standard task.
  • 01:15 This is work that we do all the time.
  • 01:18 It may be very difficult, but it's work that we know how to do.
  • 01:21 There's little uncertainty because it's our core competency.
  • 01:24 If you consider how much time it would take to accomplish one of these tasks,
  • 01:28 a distribution function of the effort needed to do the work would show a very
  • 01:32 narrow band of uncertainty.
  • 01:34 It should take somewhere between 50 hours and 60 hours,
  • 01:37 exactly where you don't know.
  • 01:40 For this type of task, select the average value as the estimate.
  • 01:44 That's the most likely case and it is the best estimate to use.
  • 01:48 Fine tuning the estimate will not be worth the effort.
  • 01:51 The next type of task is the dependent task.
  • 01:54 Once again, the activity is clearly understood, but
  • 01:57 the actual value depends upon some factor that isn't known or
  • 02:00 that can't be controlled at this point in time in the project.
  • 02:04 For instance, it may be a task that requires the use of a supplier.
  • 02:07 And depending upon which supplier is used, the amount of effort associated
  • 02:12 with doing vendor visits could be anywhere from 40 hours to 80 hours.
  • 02:16 So what should the estimate be?
  • 02:19 The 40 hour estimate, the aggressive approach, or the 80 hour estimate,
  • 02:23 the more conservative approach.
  • 02:25 I won't use something in the middle such as 60 hour estimate,
  • 02:28 which I know is wrong.
  • 02:30 What I would do in this case, is to start with the 80 hour estimate.
  • 02:34 But if the total project plan exceeds the cost or schedule constraints,
  • 02:37 I'll need to find somewhere to reduce estimates.
  • 02:40 If I need to change estimates, this task is one that I would choose to work with.
  • 02:44 I may be able to change to the 40 hour estimate right now,
  • 02:47 by getting control of that factor, that is, immediately selecting the supplier.
  • 02:53 Even if I can't control the factor, I'll switch to the aggressive estimate and
  • 02:57 then put this task in my risk register and follow it closely in my control plan.
  • 03:02 Even if I use the conservative approach in the project plan, the factor is a positive
  • 03:07 opportunity risk then and should be tracked for that reason.
  • 03:11 The third type of task is the uncertain task.
  • 03:14 With this task, we have a general understanding of the scope, but
  • 03:18 we don't know exactly what's required.
  • 03:20 The actual effort isn't clear, and that's often because it either depends on many
  • 03:24 factors we don't know or it's a type of activity we've never done before.
  • 03:28 The distribution would often look something like this,
  • 03:31 a skewed distribution.
  • 03:33 The required amount of effort could be anywhere from 40 hours to 200 hours.
  • 03:37 There are so many unknowns that an accurate estimate cannot be set.
  • 03:41 Now, I could use the best case estimate of 40 hours, but that's highly unlikely.
  • 03:47 The chances of that being correct are about the same as me winning the lottery.
  • 03:51 I could also use the worst case estimate of 200 hours.
  • 03:55 While this is the safest estimate, it's also unlikely.
  • 03:58 And if I use that approach with many uncertain tasks in the project,
  • 04:02 I will inflate my overall project timeline and budget.
  • 04:05 How about this one?
  • 04:07 I could use the most likely estimate.
  • 04:10 This is the one I normally put in my project plan that's developed with
  • 04:13 the core team members.
  • 04:15 However, when I put together the total plan,
  • 04:17 I'll use the 75 to 80% case in my totals.
  • 04:20 The difference between the most likely and the 80% is my reserve.
  • 04:25 I don't give the team the 80% when the project starts,
  • 04:27 if I give it to them, they're certain to use it all up.
  • 04:30 Instead, I give them the most likely estimate and
  • 04:33 then tell them that if they have some problems come see me.
  • 04:36 I've got some reserve to help out with that activity.
  • 04:38 So I use the most likely estimate with the team as a stretch goal.
  • 04:43 This is a high risk task which needs to go on the risk register.
  • 04:46 I'll watch this task closely because I really don't know if it's going to be 40
  • 04:51 or 200 or somewhere in between.
  • 04:54 The final type of uncertainty is a task that we don't even know about.
  • 04:58 These are sometimes referred to as the unknown unknowns.
  • 05:01 They are uncontrollable because we don't even know that the task exists in our
  • 05:04 project.
  • 05:05 On some projects, there will be activities that we have no experience with,
  • 05:09 such as a new technology or creating an application for a system we've never used,
  • 05:13 or operating in a region of the world where we have no experience.
  • 05:17 How do we estimate these tasks with any level of accuracy?
  • 05:21 You can't.
  • 05:22 I arbitrarily assign a number to a reserve.
  • 05:24 If I think there may be a lot of these, I will assign a high reserve.
  • 05:28 If I think there's not very much potential, I'll use a lower reserve.
  • 05:33 Uncertainty is a fact of life in project estimating.
  • 05:36 These are a few rules of thumb that you can use to help you with managing
  • 05:41 the uncertainty in your tasks.

Lesson notes are only available for subscribers.

PMI, PMP, CAPM and PMBOK are registered marks of the Project Management Institute, Inc.

Gift this course