# Estimating principles

The introduction to cost management explored the importance of the differences between the product and project life cycles. Understanding the total cost of the product life cycle is important in the decision to initiate the project and in making design decisions.

The project manager's main focus, however, is on developing estimates of project life cycle cost, meaning all costs related to producing the product, service, or result of the project. Estimates evolve and are refined as the project proceeds through its phases and stages.

Estimating is challenging, especially in the early stages of planning a project. As noted, cost estimating, like all of the planning processes, is iterative. Initial estimates for some tasks may have to suffice until later in the project, when more specific information becomes available.

This new information allows a more detailed estimate to be generated. The evolution of the estimate as the project unfolds mirrors the refinement of the work breakdown structure (WBS); initially, WBS components are listed at a relatively high level, and they are further decomposed as the scope is defined.

It is important for project managers to manage senior management and stakeholder expectations regarding estimates.

Far from commitments, early in the project estimates may be little more than approximations, but these approximations will become more precise as the team learns more about the details of future work from completing more elements of the definition of the project's product or service.

## What is an estimate?

An estimate is not the same as a random guess or a bid, but is the derivation of an approximate value based on one or more rational methods. An estimate uses available data for comparable activities, components or events, and extrapolates or interpolates to the current situation being estimated. In contrast, a guess is a value arrived merely by intuition.

In a limited number of instances, duration or resource requirement values may be a given. For example, a test procedure may be defined as always running for a set number of hours. Another example is that concrete takes a known amount of time to cure before construction can begin on it, based on the environmental temperature and the mixture used. In these examples, the values are givens, not estimates. Estimates are approximations.

Prior to the completion of an activity, it is not possible to know what the exact cost, duration or resource effort will be. The project manager will not be able to account accurately for some factors, such as large changes in requirements that were not factored into a project's original assumptions.

## Who is responsible for cost estimates?

In some organizations, the project manager is not responsible for developing the cost estimates. A qualified cost-estimating department may develop these instead. In other circumstances, the costs of certain types of resources, for example, in-house staff labor, are not factored into the project costs.

Within the general practice of project management it remains the responsibility of the project manager to understand and be able to justify the estimated project costs, whether or not he or she actually developed them.

## Estimating range

It is the project manager's responsibility to communicate to the project sponsor the degree of accuracy in the estimates. A single point number cannot provide sufficient understanding of the uncertainties at the beginning stages of the project. The accuracy of estimates varies significantly based on the information available at the time the estimating process is performed and the method used to generate the estimate.

Preparing cost estimates is an iterative process, performed throughout the project life cycle. As more information is discovered, the cost estimate should be revised to reflect the most current and accurate cost data available. Each time a cost estimate is revised, the project manager should review and evaluate the cost estimate to ensure that the proper technique was applied, based on the information available.

The information presented below suggests the characteristics of estimates as the project unfolds and the appropriate accuracy at that point. The subsequent sections further map these characteristics of estimates to the project phases.

### Rough Order of Magnitude (ROM)

When done: Very early in the project life cycle, during initial planning
Why done: Provides estimate of cost for selection decisions
How accurate: -25% to +75%*

Very little information is available to the cost estimator

* The percentages of accuracy indicated for these types of estimates vary within industries and application areas. The general theme is that the range of accuracy is quite broad for ROM estimates, even up to -50% to +100% in the identification phase. The accuracy range is tightened as one moves through the project to develop budgetary and definitive estimates.

### Budgetary

When done: Early, during the project definition phase as scope is defined and agreed upon
Why done: Puts dollars in the budget plans and establishes a baseline commitment to the project sponsor
How accurate: -10% to +25%

More project information is available, but changes may still occur.

### Definitive

When done: Later in the project, during project development/execution
Why done: Provides details for purchases, estimates actual costs
How accurate: -5% to +10%

Estimates are done to predict revised project completion dates/costs.

## Cost estimates over the project life cycle

### Planning estimates

Planning estimates are normally created for a proposed project before the conceptual design is completed. Planning estimates are used to set cost estimates for early project work involving scope determination and feasibility study efforts and for a preliminary budget estimate of the total project costs.

Planning estimates are based on drawing analogies from past cost experience with similar projects, where available. Rough Order of Magnitude estimates are planning estimates used in the absence of previous cost experience.

Planning estimates are typically only approximations. It is imperative that the estimator fully describe and document the basis of the estimate, including how the estimate was prepared and any items specifically excluded from the estimate.

### Budgetary or conceptual design estimates

The fundamental purposes of a budgetary or conceptual design estimate are:

• Ensuring that the project offers sufficient benefit for the estimated cost.
• Developing a reliable project cost estimate consistent with the proposed schedule.
• At a minimum, establishing baseline project costs at the phase summary level.
• Serving as the basis for project funding.

In a software engineering project, the conceptual design estimate tells the customer approximately what it will cost to complete the project within the scope defined in the project scope statement. The conceptual estimate will be based on any available conceptual design information, such as prototypes and functional specifications.

### Preliminary design or system design estimates

The preliminary design and system design estimates are intermediate estimates created for two purposes:

• Verifying that the newly available design details still support the conclusion that the revised cost budget remains less than the total project funding.

• Breaking down the cost budget to the work package level for the upcoming phase, to facilitate more complete and accurate cost monitoring and control.

Preliminary design estimates are based on the specifications that have just been developed.

Upon approval of system design and during subsequent delivery stages, system design estimates become definitive estimates. They are updated during project monitoring and control to predict revised project completion dates and costs.

### Benefits of life cycle costing

The benefits of life cycle costing include:

• Evaluating the competing options in purchasing;
• Improving the awareness of total costs;
• Providing accurate forecasting of cost profiles;
• Forecasting future resource needs;
• Supporting strategic planning and budgeting.

Fundamentally, life cycle costing is important because it helps to ensure that the least expensive alternative that will meet a project's functional requirements is selected. Initial design and development costs may constitute only a fraction of a project's overall life cycle costs.

## Estimating accuracy

The precision necessary for an estimate depends on its context. For example, the earliest estimate for a project's cost at the concept evaluation stage may be a rough order of magnitude estimate, which is good enough to determine whether the project should be pursued or not. Additional accuracy would not affect the decision, so making the effort to obtain it is not cost-effective.

On the other hand, as an upcoming stage of a project is being planned, the stage’s activities are defined at sufficient levels of detail to ensure that estimates of duration will have the necessary accuracy. The resulting project baselines will be effective for measuring performance. Without this level of detail, the project management team would not have the ability to detect and correct problems in a timely manner.

As estimates are refined further and at increasing levels of detail, the effect of diminishing returns comes into play. The project management team must determine when the estimate's accuracy is not improving sufficiently to justify further analysis.

### Influence of scope on estimating errors

Perhaps the most significant factor contributing to errors in estimates is the failure to encompass the entire technical scope. This comes primarily from two sources: failure to elicit, analyze and maintain the requirements properly, and incorrect decomposition of the work breakdown structure. The first is the reason that poor requirements are cited so often as a cause for poor project performance. The second leads to overlooked work, which often appears later as a scope change request when it was actually an error, and the scope did not change.

### Influence of assumptions on estimating errors

The assumptions about the estimate must be documented. If significant assumptions are not tested, the estimate may involve substantial error. For example, assumptions regarding productivity, resource availability, vendor capabilities, and technical feasibility should be reflected in the resource and duration estimates and spelled out in the basis of estimate (BOE) for the activity. The BOE describes how the estimates were derived and the information on which they were based.

### Influence of risk on estimating errors

When known risks are under- or overestimated, the amount of time or cost allocated to the estimate will be wrong to some degree. Experience can reduce this kind of error, but it cannot eliminate it completely.

### Influence of time on estimating errors

Estimates must accommodate the budgeting and funding cycles of the performing organization. When an estimate approaches or spans a boundary between two funding cycles, there is the inherent risk that during the project the activity may slip at least partially out of the cycle in which it was planned. This means the cost for the slipped portion will be incurred in the wrong cycle, and the available funds in both cycles will no longer match the revised cost forecast.

### Other factors affecting estimating accuracy

Some other factors that contribute to the error of an estimate include:

• Novelty of the activity and lack of experience estimating it;
• Novelty of the technology used to carry out the activity;
• Unknown skills and productivity rates of the team;
• Unexpected impediments to the progress of the activity;
• Mistakes and/or misunderstandings in the inputs to the activity;
• Changes in the inputs to the activity, particularly in the requirements;
• Uncontrolled changes in working environment that impair productivity;
• Effects of geographic location of work;
• Locations of team members relative to the work;
• Available methods of team communication;
• Quality of support services;
• Availability of appropriate tools; and
• Availability of subject matter experts.

## Cost drivers

Cost drivers are project cost elements that heavily impact the overall project costs. They will either inflate or moderate the base cost estimate that resulted from the measure of the project's scope. Small changes in cost drivers result in large changes in the overall project costs.

Parametric cost estimating relies on an understanding of these cost drivers. They result in additional estimating measures that must be used to adjust the cost estimates upward or downward to account for these factors.

Project managers should be concerned about cost drivers because they represent specific elements of the project that could potentially become budget busters. Cost drivers are the areas worth spending extra time on when developing the estimate in order to get more accurate data on quantity required and unit costs.

When reviewing a cost estimate, recognizing and focusing on cost drivers is more productive than trying to accurately forecast cost for all project elements.

## Environmental factors

The cost estimating process requires the project management team to consider two environmental factors:

• Marketplace conditions - This involves the products, services, and results that are available in the marketplace, their sources, and the terms and conditions of their use. These conditions will affect the choice of resources, and their prices will then serve to establish their cost estimates. In addition, if fluctuations in these conditions are expected, prices may change during the project. These price changes will also affect the cost estimates.
• Commercial databases - Resource cost rates are often available from commercial databases that track skills and human resource costs and standard costs for material and equipment. With adjustments for local variation, these may serve as the base prices for the cost estimates.

## Sources of information

Organizational assets that should be considered when estimating include:

• Cost estimating policies - If the organization has established mandatory approaches for cost estimating in various circumstances, the project management team must comply with them;

• Cost estimating templates - Often the finance function or previous project teams have developed templates that are based on previous projects. These templates can be reused and tailored for use on new projects to suit their characteristics;

• Historical information - Often the performing organization maintains a repository of information about work that has been performed in the past that was similar, or that used the same resources planned for the current project. In this case, there may be available historical information that the project management team can leverage in developing the cost estimates;

• Project files -Records from prior projects that have performed similar work can help with developing cost estimates;

• Team knowledge - Frequently the team members will have been selected because of their previous experience with similar work. This usually provides a base of knowledge about similar efforts that can be used to guide cost estimates;

• Lessons learned - If previous cost estimates are available from a previous project that was similar in scope and size, there will often be lessons learned about how to (and how not to) generate and tailor the estimates to be even more accurate than those used on the previous project.

Thanks to Ignacio Manzanera for providing this book