Estimating Completion versus Level of Effort


Estimating "Completion" versus "Level of Effort"

In almost every project there are some WBS elements for which the tasks and activities of a deliverable cannot be scoped with certainty. Estimates for cost accounts of this type are called "level of effort." Level of effort describes a concept of indefinite scope and application of "best effort" by the provider. How then to contain the risk of the estimate? We call on three-point estimates, a sober look at the most pessimistic possibilities, and the use of statistical estimates to get a handle on the range of outcomes.

Completion estimates are definitive in scope, though perhaps uncertain in total cost or schedule. After all, even with a specific scope there are risks. Nevertheless, completion estimates have a specific inclusion of tasks, a most likely estimate of resource requirement, and an expectation of a specific and measurable deliverable at the end. In the examples presented in this chapter, the underlying concept is completion.

Let us consider a couple of examples where level of effort is appropriate. Project management itself is usually assigned to a WBS cost account just for management tasks. Although there are tangible outcomes of project management, like plans, schedules, and budgets, the fact is that the only real evidence of successful completion of the cost account is successful completion of the project. Work packages and cost accounts of this type are usually estimated from parametric models and "similar-to" estimates, but the total scope is indefinite.

Research and development efforts for "new to the world" discoveries are an obvious case for level of effort, particularly where the root problem is vague or unknown or where the final outcome is itself an "ah-hah!" and not specifically known in advance. In projects of this type, it is appropriate to base funding on an allocation of total resources available to the business, set somewhat short-term milestones for intermediate results, and base the WBS on level of effort tasks. If prior experience can be used to establish parameters to guide the estimating, then that is all to the advantage of the project sponsor and the project manager.



Summary of Important Points

Table 3-9 provides the highlights of this chapter.

Table 3-9: Summary of Important Points

Point of Discussion

Summary of Ideas Presented

Work definition and scoping

  • Organizing and defining the project scope of work is a required prerequisite to all analytical estimates concerning projects.

  • The WBS serves two purposes: (1) organizing the work and (2) providing data on the project deliverables.

  • The WBS is not a project organization chart or a project schedule.

  • The WBS may support more than one view of the project: sponsor's view, developer's view, operations and maintenance view.

  • Most often, the WBS is a hierarchical structure of the project deliverables.

The OBS and RAM

  • The people who work on projects may be assigned exclusively to the project or participate in the project and a "home" organization. The "home" organization breakdown structure is called the OBS.

  • The OBS maps to the WBS by means of the RAM.

Work packages and cost accounts

  • The work package is the lowest level of work identified with cost on the WBS.

  • The cost account is a summation of subordinate work packages.

Chart of accounts

  • The chart of accounts is a WBS of the business organized by financial accounts like expense accounts, capital accounts, and equity accounts.

  • The WBS is typically an extension of the chart of accounts structure, touching expense and capital accounts most often.

WBS dictionary

  • The WBS dictionary of a set of definitions of the work packages and cost accounts on the WBS.

WBS baseline

  • The WBS baseline is the scope and cost decided on at the outset of the contract. The baseline is managed as a fixed set of numbers until changed through a change management process.

Project estimates

  • The objectives of performing an estimate are twofold: to arrive at an expected value for the item being estimated and to be able to convey a figure of merit for that estimate.

Top-down estimates

  • Top-down value judgments from the business side of the project balance sheet conveyed to the project team.

Similar-to estimates

  • Similar-to judgments from either side of the project balance sheet, but most often from the other business side.

Bottom-up estimates

  • Bottom-up facts-driven estimates of actual work effort from the project side of the project balance sheet conveyed to the project sponsor.

Parametric estimates

  • Parametric calculations from a cost-history model, developed by the project team, and conveyed to the project sponsor.

Confidence intervals in estimates

  • Every estimate should include some assessment of the risk of completing the project for the estimated figures.

  • The Normal distribution is invoked by means of the Central Limit Theorem to describe the project outcome distribution probabilities.

  • Cumulative probabilities from statistical distributions provide the data to assess the confidence of meeting the project outcome.

  • Confidence intervals can be estimated for any WBS regardless of the estimating methodology.

Completion and level of effort estimates

  • Level of effort is appropriate where the scope is indefinite.

  • Completion should be used whenever it is possible to make the scope definitive.