When it comes to project planning, it is crucial to consider a number of key factors, including work effort, duration, costs, people, and physical resources. Estimates play a significant role in this process, serving as a quantitative assessment of what's likely to happen in your project. As the project progresses, these estimates are subject to change, depending on current information and circumstances - and progressive iteration. Four aspects of estimating that are impacted by the project's phase in its life cycle are range, accuracy, precision, and confidence.
Estimates are a quantitative assessment of the likely amount or outcome of a variable, such as project costs, resources, effort, or duration. They are based on the information available at the time they are made and can change as the project evolves and more information becomes available. It is important to consider adjusting estimates for uncertainty, as estimates are inherently uncertain and are associated with risk. To mitigate this risk, contingency time or funds may be added based on the results of a simulation conducted to establish the range of uncertainty.
The Range of Estimates
At the start of a project, when there's limited information available, estimates tend to have a broad range. This range narrows as the project evolves and more information becomes available. For example, projects that are well into their life cycle may have an estimating range of minus 5 to plus 10 percent.
Accuracy and Precision
Accuracy refers to the correctness of an estimate. It is closely linked to the range of the estimate. The higher the accuracy, the closer the prediction is to the actual value. On the other hand, precision refers to the degree of exactness associated with the estimate. A precise estimate, such as "2 days," is more precise than a vague estimate, such as "sometime this week." To ensure the precision of an estimate is compatible with the desired accuracy, it is important to consider both accuracy and precision when estimating.
Confidence in Estimates
Estimates are inherently uncertain, and uncertainty is associated with risk. Therefore, it is important to gather enough data to have confidence in your estimates. Experience plays a significant role in building confidence in estimates. If you have previous experience working on a similar project, you can be more confident in your estimates. On the other hand, if you're working on a new project with evolving technology components, your confidence in your estimates is likely to be low. Key deliverable dates or budget estimates may be adjusted, or contingency time or funds may be added, based on the outcomes of a simulation to establish the range of uncertainty for these parameters.
Deterministic and Probabilistic Estimating
Deterministic estimates, also known as point estimates, present a single number or amount, such as 36 months. Probabilistic estimates, on the other hand, include a range of estimates along with the associated probabilities within that range. They can be developed manually by creating a weighted average based on multiple likely outcomes or by running a simulation to develop a probability analysis. These estimates can be developed through weighted averages or computer simulations to develop a probability analysis of a particular outcome. Probabilistic estimates derived from computer simulations have three key components: a point estimate with a range (e.g., 36 months +3 months/-1 month, a statement of confidence (e.g., a 95% confidence level), and a probability distribution (e.g., the dispersion of the data within and around the given range - such as a standard deviation of X days). Together, these three components form a complete metric for a probabilistic estimate.
Absolute and Relative Estimating
Absolute estimates are specific, using actual numbers, while relative estimates are presented in comparison to other estimates. Relative estimates are shown in comparison to other estimates and only have meaning within a given context. For example, an absolute estimate for effort might be 120 hours of work, while a relative estimate might be compared to the effort required for previous work.
Flow-based Estimating
Flow-based estimates are developed by determining the cycle time (the total elapsed time it takes one unit to go through a process) and throughput (the number of items that can complete a process in a given amount of time) of a process. This provides an estimate of how long it will take to complete a specified amount of work. These two numbers can provide an estimate for completing a specified amount of work.
Key concepts:
- Planning is the process of developing estimates for work effort, duration, costs, people, and physical resources.
- Estimates are quantitative assessments of the likely amount or outcome of a variable, such as project costs, resources, effort, or duration.
- Range is the broadness of an estimate, which tends to be wider at the start of a project and narrows as the project evolves.
- Accuracy is the correctness of an estimate.
- Precision is the degree of exactness associated with an estimate.
- Confidence is the level of certainty in an estimate, which increases with experience.
- Deterministic Estimating is a method of presenting a single number or amount as an estimate, such as 36 months.
- Probabilistic Estimating is a method of presenting a range of estimates along with the associated probabilities within the range.
- Absolute Estimating is a method of presenting specific information and actual numbers as an estimate, such as 120 hours of work.
- Relative Estimating is a method of presenting estimates in comparison to other estimates, only having meaning within a given context.
- Flow-based Estimating is a method of determining an estimate by calculating the cycle time and throughput.
- Adjusting Estimates for Uncertainty is a process of modifying estimates to account for inherent uncertainty and risk.
- Point Estimate is a single number or amount as part of a probabilistic estimate.
- Confidence Level is a statement of confidence as part of a probabilistic estimate, such as a 95% confidence level.
- Probability Distribution is a description of the dispersion of data within and around a range as part of a probabilistic estimate.
- Planning Poker is a form of relative estimating where a project team comes to a consensus on the effort necessary to deliver value.
- Story Points are units of measurement used in planning poker to estimate work effort.