How to use parametric estimating to improve project performance
Project estimation is a challenging task, with numerous variables and little room for error.
However, parametric estimating is a data-driven technique that can help project managers
accurately estimate the cost, time, and resources required for a project.
In this guide, we will explore the world of parametric estimating in-depth, covering the critical
concepts, tools, and techniques needed for project managers to get the most out of this technique.
What is parametric estimating?
Parametric estimating is a powerful technique project managers use to quickly and accurately
estimate the cost, time, and resources required for a project using various data. It has become
increasingly popular as organizations try to manage projects more efficiently and effectively.
Parametric estimating relies on comprehensive regression analysis, a statistical technique that
examines the relationships between different variables to identify patterns and trends. By
analyzing historical data on similar projects, regression analysis can help project managers
identify the factors that influence project parameters and develop predictive models to estimate
these parameters in new projects. We’ll go into more detail later on. But first — let’s define
some key terms.
Key terms you need to know
• Task or part: A task or part is a discrete component of a project that can be estimated
using parametric estimating. Project management of a construction job might call for
installing electrical wiring or pouring concrete for a foundation. This would count as a
task or part; it’s a minor job that goes into completing the overall thing.
• Parameter: A parameter is a characteristic or factor influencing a task or part’s time,
cost, or resource requirements. Examples of parameters include the size of a building, the
number of workers required for a task, or the complexity of a software application.
• Cost per parameter unit: The cost per parameter unit is the amount of money required
to complete a task or part for a given parameter value. For example, the cost per square
foot of construction might vary depending on the type of building.
• Time per parameter unit: The time per parameter unit is the time required to complete
a task or part for a given parameter value. For example, the time required to install
electrical wiring might vary depending on the size of the building.
• Parameter value: The parameter value is the value of a parameter for a given task or
part. For example, the parameter value for the size of a building might be expressed in
square feet.
You also need to know two key concepts: deterministic and probabilistic estimates.
Deterministic vs probabilistic estimates
Deterministic estimates are based on a single, fixed value for each estimated parameter. This
means that the estimated value for each parameter is assumed to be accurate and will not vary
during the project.
Deterministic estimates are helpful when there is a high degree of certainty and predictability in
the project parameters. Examples of deterministic estimates include:
1. A construction project estimate that assumes a fixed cost per square foot of building.
2. An engineering project estimate that assumes a fixed time to complete a design phase
based on historical data.
On the other hand, probabilistic estimates consider the probability or likelihood of a range of
values for each parameter being estimated. This means that the estimated value for each
parameter is not assumed to be accurate and may vary during the project.
Probabilistic estimates are helpful when there is high uncertainty or variability in the project
parameters. Examples of probabilistic estimates include:
1. A software development project estimate that considers a range of possible time and cost
estimates for each development task based on historical data and expert judgment.
2. A manufacturing project estimate that considers the probability of defects or quality
issues during the production process and incorporates a contingency plan to account for
these potential issues.
Where parametric estimating fits in with the four estimating techniques
There are four main techniques for estimating project parameters: parametric estimating, bottom-
up estimating, analogous estimating, and three-point estimation. Each method has its strengths
and weaknesses, and project managers may choose to use one or a combination of these
techniques depending on the project and its requirements.
Parametric estimating falls under the category of top-down estimation methods and analogous
estimating. Top-down estimation methods rely on data from previous projects or expert
judgment to estimate the parameters of a new project — a useful approach when historical data is
available and when there is a high degree of similarity between the current project and previous
projects.
Parametric estimating is particularly suited to projects with well-defined parameters, such as the
number of units for production. Using statistical models and historical data to estimate the cost
and time required for each unit of work, project managers can quickly develop accurate estimates
for the entire project.
What kinds of projects benefit from parametric estimating?
Parametric estimating is a valuable technique for a range of projects, but as we mentioned above,
it works best for those with set parameters. Here are some examples to illustrate the point:
1. Construction Projects: Projects often involve repetitive tasks like laying bricks or
pouring concrete. Project managers can quickly develop highly accurate estimates for the
entire project using historical data and statistical models to estimate the time and cost
required for each task.
2. Manufacturing Projects: Similar to construction projects, manufacturing projects often
involve the production of large quantities of standardized products. By estimating the
cost and time required for each production unit, project managers can develop accurate
figures for the whole job.
3. Software Development Projects: Projects often involve many discrete tasks, such as
coding, testing, and debugging. Looking at past job data can give project managers fairly
accurate insights into future projects.
4. Engineering Projects: Projects often involve designing and developing complex systems
like bridges or airplanes. By breaking down the project into its smallest components and
estimating the time and cost required for each component, project managers can develop
accurate estimates for the entire project.
It’s important to note that while parametric estimating can be a valuable technique, it’s not
always the best approach. Projects with high uncertainty or variability may be better suited to
other estimation techniques, such as three-point estimation or expert judgment.
How to do parametric estimating
Here’s a step-by-step guide to help you get the most out of this technique.
Step 1: Define its use in the project scope
First, you must determine which parts of your project you can estimate using this method. This
involves breaking your project into smaller, more manageable parts or tasks.
Creating a work breakdown structure (WBS) is an excellent place to start. To do this, review
your project scope and identify the major deliverables. Then, break down each deliverable or
outcome into smaller tasks or components. Each task should be defined in terms of deliverables
and resources needed.
Identify the tasks or parts of the project that need to be estimated and ensure all stakeholders
agree.
Step 2: Gather historical and market data
Next, you need to gather information on past projects similar in scope, scale, and complexity to
the tasks you identified in Step 1.
Start by identifying the characteristics of past projects that are most relevant to the tasks you’re
estimating. For example, if you’re working on the cost and time required for a construction
project, you might look for past projects similar in size, complexity, and location.
Once you’ve identified the most important factors, gather data on these similar projects’ cost and
time requirements.
When gathering data, it is essential to consider the relevance and reliability of the information
you’re collecting. Data on historic projects may not directly apply to your project if there have
been significant changes in technology, regulations, or market conditions. Similarly, market
trend data may not be reliable if it is based on a small sample size or doesn’t account for regional
or other variations.
As well as delving into your own data, this stage might involve researching publicly available
information, such as project reports or industry publications, or contacting industry experts or
professional associations for information and advice. Don’t forget to use your competitors’
market information, too.
Step 3: Identify parameters
Next, you identify the data you want to test for correlation with the cost or time values. By
testing for correlation, you can develop a model to estimate future projects’ cost and time
requirements based on similar parameters.
Start by reviewing the data you’ve collected. Look for commonalities, such as the number of
workers, the amount of materials used, or the duration of the project. You may also want to
consider other factors that could impact the cost and time requirements of the project, such as
weather conditions or regulatory requirements.
Once you have identified potential parameters, you must test them for correlation with the cost or
time values. This involves using statistical methods (more on this later) to determine whether
there is a significant relationship between the parameters and the cost or time values. You’ll need
a regression analysis tool to do this.
Once you have identified the parameters most strongly correlated with cost or duration, you can
determine which drives these two things the most. And from there, you can begin thinking about
the various tasks and their impact.
It’s important to note that some parameters may strongly correlate with cost or duration
but may not be driving the cost or duration. You may need further statistical analysis to
identify the true drivers in these instances. For example, a project with a longer duration also has
a higher cost. In that case, it may be challenging to determine whether the longer duration drives
the cost or if other factors are involved.
Regression analysis tools: what are they?
Regression analysis tools use statistical techniques to analyze the data and determine the strength
of the relationship between the parameters and the cost or time values. These tools typically
provide graphical representations of the data, including scatterplots and regression lines, as well
as statistical outputs such as R-squared values, p-values, and confidence intervals.
Examples of comprehensive regression analysis tools include Microsoft Excel’s built-in
regression analysis tool, SPSS, SAS, and R. These allow users to input data and perform
statistical analyses to identify the parameters that significantly impact the cost or time
requirements of the project. (Microsoft has a great guide on using its statistical analysis tool.)
Step 4: Develop a model and run backrests
For more complex estimates or projects, developing a model that incorporates the identified
parameters and their correlations with cost or duration is helpful. You can then use this model to
make more accurate estimates later on.
Perform backtesting to ensure your model’s accuracy. Backtesting involves applying the model
to historical projects and comparing the estimated costs or durations to actual ones. This allows
you to determine how well the model performs and identify areas where it may need refinement
or adjustment.
Backtesting also helps you identify any biases or errors in the model, such as assumptions about
the relationship between parameters that do not hold in reality. By identifying and correcting
these issues, you can improve the accuracy of your estimates and reduce the risk of cost or time
overruns.
Remember to continually monitor and update the model as new data becomes available or as the
project evolves. This allows you to adjust your estimates as needed and ensure they remain
accurate and reliable throughout the project.
Step 6: Calculate your estimate
Once you’ve identified the parameters that drive cost or duration, use them to compute your
parametric estimate(s). There are a few ways to do this, but we will use the Cost Estimating
Relationship (CER) formula.
The CER formula is expressed like this:
E_parametric = A_old / P_old x P_curr
• E_parametric: This is the parametric estimate, which is the estimated cost or duration of
the current project.
• A_old: This is the historical amount of cost or time for a similar project or task.
• P_old: This is the historical value of the parameter (cost driver) for the similar project or
task.
• P_curr: This is the value of the parameter (cost driver) for the current project.
Two real-world examples
Let’s take a look at the formula in action.
Example 1: Construction
Suppose a construction company estimates the cost and duration of building a new bridge based
on previous construction projects. They have historical data showing that the average cost of
building a bridge is $10 million, and the average time is two years. They also know that the
length of the bridge is the most significant factor affecting the cost and duration. The length of
the previous bridges ranged from 500 feet to 2000 feet, with an average of 1000 feet.
Using the parametric estimating formula, the company can estimate the cost and duration of the
new bridge as follows:
E_parametric = $10,000,000 / 1000 feet x 1500 feet = $15,000,000
Therefore, the company can estimate the cost of building the new 1500-foot bridge will be $15
million.
Example 2: Software development
Suppose a software development company estimates the cost and duration of developing a new
mobile app based on previous app development projects. They have historical data showing that
the average cost of developing an app is $100,000, and the average time is six months. They also
know that the number of features in the app is the most significant factor affecting the cost and
duration. The previous apps ranged from 10 to 50 features, with an average of 30 features.
Using the parametric estimating formula, the company can estimate the cost and duration of the
new app as follows:
E_parametric = $100,000 / 30 features x 40 features = $133,333
Therefore, the company can estimate the cost of developing the new 40-feature app will be
$133,333.
With this formula, you can work out your estimated figure. However, it’s important to note that
the estimate’s accuracy depends on the quality and relevance of the historical data used, so use it
with this in mind.
Advantages and disadvantages of parametric estimation
Advantages
First, the advantages.
1. Saves time and effort: With parametric estimation, the estimator saves time by using
historical data and statistical techniques to make estimates. They don’t have to start from
scratch.
2. Improves accuracy: This method is more accurate than other estimation forms, such as
analogical or bottom-up, because parametric estimation considers more variables,
reducing the margin of error.
3. Provides consistency: It gives a consistent method for estimating costs and time. This
consistency means multiple estimators can produce similar results, reducing the variation
risk.
4. Enables data-driven decisions: Basing decisions on historical data and statistical
analysis helps project managers make objective decisions based on data rather than
relying on subjective opinions.
5. Facilitates continuous improvement: By tracking and analyzing the actual costs and
time spent on projects, parametric estimation provides valuable feedback for future
projects. You can then use this feedback to improve future estimates, increasing accuracy
and efficiency over time.
6. Enables scenario analysis: Project managers can simulate different scenarios to
determine the impact of changing parameters on cost and time, which helps them make
informed decisions regarding resource allocation and risk management.
Disadvantages
Next, let us take a look at the disadvantages.
1. Lack of accuracy: It relies on historical data and the assumption that the relationship
between the parameters and the cost or duration of a project remains constant. But this
isn’t always the case; the complexity of a project and the availability of resources can
vary. (To mitigate this, use relevant historical data, and consider any differences during
the estimation process.)
2. Limited scope: Parametric estimating is suitable for projects with well-defined
parameters. It might not always be practical for complex projects that require subjective
judgments or that are difficult to quantify. (To overcome this, make sure that the
parameters used in the estimation process are relevant and comprehensive.)
3. Requirement for historical data: Parametric estimating relies on historical data to
estimate project costs and durations. This approach won’t work if this data is unavailable
or outdated.
4. Limited flexibility: Assuming that the relationship between the parameters and the cost
or duration of a project remains constant limits the flexibility of the estimation process,
especially when there are changes to the project scope or other parameters. (To address
this, review your estimation process regularly.)
5. Lack of understanding: Parametric estimating requires significant technical expertise
and knowledge of statistical analysis, making it challenging for individuals who do not
have a background in these areas to use the technique effectively. (If you opt for this
approach, provide training and support to those responsible for estimating project costs.)
Project management software: your secret weapon
Project management software can be crucial in facilitating parametric estimation and improving
accuracy. Here’s how:
1. Historical data analysis: Project management software can help collect and analyze
historical data from previous projects. It also helps identify the factors influencing these
parameters and how they will likely impact the current project.
2. Parametric modeling: Project management software can assist in developing, building,
and updating parametric models for estimating project parameters.
3. Collaboration: Project management software can facilitate collaboration among team
members, stakeholders, and subject matter experts, making it easier to incorporate
insights and feedback into the estimation process. With Backlog, our project management
tool, you have the bonus of real-time updates, ensuring everyone is on the same page.
4. Scenario planning: Project management software can help conduct scenario planning
and what-if analysis, which might involve changing the assumptions and inputs in the
parametric models to simulate different scenarios. You can then test various risks and
uncertainties and develop a contingency plan.
5. Tracking and monitoring: With project management software, you can track and
monitor project progress and performance and compare actual results to your estimations.
The software can also trigger notifications when the parameters exceed the expected
range, allowing project managers to take corrective action faster.
By leveraging the software’s capabilities, project managers can make more accurate estimates,
identify risks and uncertainties, and develop contingency plans to mitigate them, leading to a
safer, more precise project.
