What is three-point estimation?

Three-point or multipoint estimation is a technique used in project management and other disciplines to estimate future events, typically costs or task durations. It incorporates uncertainty and risk by considering three scenarios: optimistic, most likely, and pessimistic. The approach usually involves two models: the Triangular Distribution, which gives equal weight to all estimates, and the PERT (Program Evaluation and Review Technique) distribution, which emphasizes the most likely estimate. This method helps project managers to make more informed decisions by providing a more realistic range of outcomes than a single-point estimate.

How to perform multipoint estimation for task duration

Get your project team together and ask them to estimate what the duration of each task would be under each of these circumstances:

1. Everything goes well - no delays or mistakes (Optimistic estimate).
2. What will probably happen - assumes there will be some issues (Most Likely estimate.
3. Everything goes wrong - everything that can go wrong does (Pessimistic estimate).

Record all the estimates and use them to calculate PERT and Triangular Distribution.

How to Calculate PERT (Program Evaluation and Review Technique)

The PERT Distribution formula places more emphasis on the most likely estimate, using a weighted average:

Average = \( \frac{{\text{Optimistic} + 4 \times \text{Most Likely} + \text{Pessimistic}}}{6} \)

The most likely estimate is multipied by 4 reflecting the belief that the most likely scenario is more probable than the extreme cases. It is based on probability theory and statistics, specifically Beta Distribution (Project Management Academy, 2021). The division by "6" effectively considers the spread over six equal segments between the optimistic and pessimistic estimates. This division by six also translates into the standard deviation calculation, reflecting the spread across the range in a way that suits the beta distribution typically used in PERT analysis.

How to Calculate Triangular Distribution

In three-point estimation, the Triangular Distribution formula calculates a simple average of the three estimates:

Average = \( \frac{{\text{Optimistic} + \text{Most Likely} + \text{Pessimistic}}}{3} \)

This treats all estimates as equally likely.

What is Standard Deviation?

Standard deviation is a statistical measure that quantifies the amount of variation or dispersion in a set of values. In project estimating, standard deviation is used to assess the risk associated with the estimates provided. It measures how much the actual outcomes are expected to deviate from the average estimate. This is crucial for project managers because it helps them understand the uncertainty involved in time and cost estimates, enabling them to plan buffers and make informed decisions to manage potential risks in project timelines and budgets.

How to calculate Standard Deviation for PERT

Standard Deviation (\( \sigma \)) = \( \frac{{\text{Pessimistic} - \text{Optimistic}}}{6} \)

The symbol stands for Sigma. In the context of statistics and project estimation, sigma (\( \sigma \)) typically represents the standard deviation of a set of values. It measures the amount of variation or dispersion from the average (mean). In project management, particularly when using PERT, sigma gives an idea of the uncertainty around the estimate, with a higher sigma indicating greater uncertainty and variability.

Does standard deviation apply to PERT or Triangular distribution?

Standard deviation is typically calculated for the PERT estimate in project management, not for the Triangular result. The PERT estimate uses a weighted average that emphasizes the most likely outcome, making it more informative to measure variability around this estimate with standard deviation. The Triangular result, being a simple average, does not typically involve standard deviation as it does not weigh the estimates differently, thus providing less insight into the likelihood or risk of variances.

How to use Three-Point Estimation on your Project - a worked example for a mobile app

Lets imagine you are building a mobile app. You know what you want it to do and look like, but you don't have clear requirements other than that. You get your team together and estimate the number of days based on expert input from your developer and previous project data provided by your project manager. You agree on the following estimates and enter them in the calculator:

• Optimistic - 100 days
• Most likely - 120 days
• Pessimistic - 200 days

PERT Estimate
130 days = \( \frac{100 + 4 \times 120 + 200}{6} \)

Triangular Estimate
140 days = \( \frac{100 + 120 + 200}{3} \) days

standard Deviation
16.67 \( \sigma = \frac{{200 - 100}}{6} \) days

The results are:

• PERT - 130 days
• Triangular - 140 days
• Standard deviation - 16.67 days

What do the results mean?

The PERT estimate at 130 days suggests that, considering the weighted approach (emphasizing the most likely duration), the project is expected to be completed around this time. The Triangular estimate being higher at 140 days means that when giving equal weight to all scenarios, the average is skewed towards the longer duration. A standard deviation of 16.67 days indicates a relatively wide spread around the PERT estimate, suggesting significant uncertainty or variability in the estimates provided. This could imply a need for a contingency or buffer in the schedule to account for potential deviations from the PERT estimate.

Deciding on the best estimate to use

You can decide on the best estimate to use by considering a few key factors:

1. Project Complexity and Uncertainty: For projects with high uncertainty or variability in task durations, the PERT estimate may be more appropriate because it considers this variability by giving more weight to the most likely outcome.
2. Risk Tolerance: If the project or stakeholder has risk tolerance, using the PERT estimate with its associated standard deviation can help in setting more conservative timelines and budgets.
3. Historical Data and Expertise: If past project data or expert insights suggest that certain tasks tend to lean towards optimistic or pessimistic outcomes, this can influence whether the triangular or PERT estimate might be more realistic.
4. Criticality of Task: For critical tasks, it might be safer to consider the PERT estimate and its standard deviation to ensure that potential delays do not impact overall project delivery.

What estimate might be chosen for the mobile app project?

In this scenario given the lack of clear requirements but availability of expert input and historical data, I might lean towards using the PERT estimate for building the mobile app with between 5 to 15% contingency. For several reasons:

• Expert Judgment: Incorporating insights from the developer, who has firsthand knowledge of similar tasks, provides a more accurate estimate.
• Historical Data: Previous project data can help predict future performance and timelines, enhancing the reliability of the PERT estimate.
• Handling Uncertainty: The PERT method, by emphasizing the most likely scenario and calculating standard deviation, helps in managing the uncertainties associated with the project, ensuring a buffer is included for unforeseen delays.

What are the advantages of multipoint estimating?

Three-point or multipoint estimating offers several advantages:

• Improved Accuracy: By considering optimistic, most likely, and pessimistic scenarios, it provides a more realistic range of possible outcomes compared to single-point estimates.
• Risk Management: This method helps identify potential risks by highlighting the variance between the optimistic and pessimistic scenarios, allowing for better contingency planning.
• Increased Confidence: Stakeholders may have more confidence in estimates that account for different scenarios, improving decision-making and project planning.
• Flexibility: It allows project managers to adjust plans and expectations based on a spectrum of potential outcomes, not just a single estimate.

These benefits make three-point estimating particularly valuable in complex projects where uncertainty is high.

What are the disadvantages?

Using multipoint estimation, such as the three-point method, has some disadvantages:

• Subjectivity: The estimates (optimistic, most likely, pessimistic) rely heavily on the judgment of those providing them, which can introduce bias and inconsistency.
• Complexity: It can be more complex and time-consuming than single-point estimates, requiring more data and analysis.
• Reource intensive: It often means involving multiple experts or teams, which can consume significant time and resources. This can make it less practical for smaller projects or when quick decisions are needed.
• Misunderstanding of Outputs: The results, such as triangular distribution and standard deviation, might be misunderstood or misapplied by stakeholders, leading to incorrect expectations.

These factors can make multipoint estimation challenging to use effectively in some project scenarios.

What projects suit multipoint estimating?

Multipoint estimating is particularly well-suited for projects where:

1. Complexity and Uncertainty are High: Large or complex projects with many unknown variables benefit from this method as it helps gauge risks and manage expectations.
2. Innovative or Unique Endeavors: Projects that involve new technologies or processes, where historical data may be limited, can use expert judgment in a structured way through multipoint estimating.
3. Strategic or High-Impact Projects: For projects where the stakes are high, such as in strategic business initiatives, this method helps in understanding the potential scope of outcomes and preparing for various scenarios.

4 real-world examples

1. Software Development Projects: Especially for new product developments where technology or user requirements may evolve during the project.
2. Research and Development: Projects involving innovative products or processes where outcomes are highly uncertain.
3. Pharmaceutical Trials: Where the duration and outcomes can vary widely based on regulatory feedback and clinical results.
4. Infrastructure Projects: Large scale constructions like highways or dams, where environmental, regulatory, and technical risks can significantly impact schedules.

These examples involve high variability and risk, making multipoint estimating a useful tool for planning and decision-making.

What project don't suit multipoint estimating?

Multipoint estimating might not be suitable for projects where:

1. Simple or Routine Tasks: Projects involving straightforward tasks with predictable outcomes, where complexity and uncertainty are minimal.
2. Low-Risk Projects: Smaller projects with fewer variables and lower stakes, where extensive risk assessment and variability analysis may not be necessary or cost-effective.
3. Tight Deadlines: Projects with severe time constraints may not have the luxury to delve into detailed scenario-based estimating, requiring faster, less detailed estimation methods.

4 real-world examples of projects that are more suited to single point estimating

1. Small Construction or Renovation Projects: Such as installing a new roof on a house or updating a bathroom, where the process is straightforward and the contractor has extensive experience with similar tasks.
2. House Building: the development of houses following standard design, quantities and materials. For example, houses built by the UK's biggest house builder, Barratt Developments.
3. Event Planning: For events that follow a standard format and scale, like a regular annual conference, where previous editions provide a clear template for planning and execution.
4. Software Installation and Configuration: For standard software setups in a business environment, where the steps and time required are well-established and predictable based on past installations.

These projects generally have clear, well-defined tasks and outcomes, making them suitable for single-point estimates where variability is minimal.

References

Project Management Academy A Three-Point Estimating Technique: PERT Accessed: 27 April 2024.