When it comes to project planning, estimation is essential for generating precise estimates of your project costs. Otherwise, your project's budget may be less than your actual costs, or overly excessive in some instances. There are several project estimation techniques used by project managers when estimating costs, one of which is parametric estimating.
Parametric estimating has a number of distinct advantages for project managers who want to be confident in their new project costings.
Parametric estimating is quantitative and uses statistics to calculate the expected amount of resources needed to complete your project, whether it be cost or time, or even human resources. Project managers use parameters or characteristics based on historical data or past projects when calculating estimates.
Because this method uses adequate historical data, parametric estimating can provide accurate and precise estimations. Project managers consider parametric estimating as thought-through and reliable for formulating project estimates and deciding on a project budget.
In parametric estimating, you can produce estimates of varying levels of detail through different approaches. The Project Management Institute highlights two results you can calculate with parametric estimating:
Deterministic estimates are calculated based on parametric scaling. This is a single number that shows the expected resources needed for your current project, usually either costs or time. If you know there are differences between your current project and the previous project that you used for the estimates, you can adjust the estimate manually.
Rather than a single result, probabilistic estimates show a range of estimates based on varying costs and durations. Project managers use a probability density curve with three benchmarks when presenting probabilistic estimates:
When calculating probabilistic parametric estimating, you calculate the optimistic and pessimistic estimates through a target probability. You can then derive a final estimate from the three estimates using a similar approach as a triangular or PERT distribution.
There are other techniques to estimate project costs. One of these is the analogous estimating technique. Like parametric estimation, this technique can help generate project cost and duration estimates.
Also considered a top-down estimating technique, analogous estimation uses data from a similar past project to estimate costs, duration, and resources needed for the current project.
The method takes its name from "analogy" because the analogous estimate technique relies on the similarities and parallels between your the previous projects and your new venture. Accurate estimations using this estimating process depends on having reliable data points.
Both parametric estimating and analogous estimates use historical data to calculate project cost estimates. However, these two methods differ in that parametric estimating uses statistical methods and publicly available data to calculate the estimates, while analogous estimating draws estimates from your own similar past projects.
While both require historical information, one requires calculation and statistics, while the other requires expert judgment. It takes a lot of time to perform parametric estimating, but the result is more detailed and accurate than other estimating techniques.
Despite its accuracy and precision, parametric estimating is not always used for estimating project values. There are three conditions that need to be met every time you use parametric estimation. If one of these conditions isn't present, you may experience difficulty in calculating your estimates with this technique, and you should use other techniques, such as analogous estimation.
The three conditions are:
A project manager cannot perform parametric estimations without sufficient historical data on each of the parameters involved in the whole project. You need to collect information from various sources if you wish to engage in parametric estimation.
Since parametric estimation is a quantitative technique, parameters must be quantifiable. You need to be able to convert your data into each parameter value in recognizable units, such as hours and dollars.
Parametric estimating applies quantitative methods to determine the statistical relationship between parameter values. This means that these parameters must be scalable so that accurate predictions can be made, even if there are differences in the projects that are being compared.
For example, if a previous construction project involved three hundred people-days to complete, the parametric estimating technique only works if the project team is scalable. The estimated values for labor costs on the current construction project will be extrapolated from the underlying data.
If, instead, the project team is not scalable, and you would look at alternative solutions for the project execution (such as outsourcing certain aspects), parametric estimation is not a suitable technique for estimating cost.
Like any other quantitative method, parametric estimations undergo a process. The steps are similar, but differences can exist depending on the size and complexity of the project. The Project Management Institute has identified five steps.
When using quantitative methods, it is essential that you first determine the parts you can apply to your chosen method. You should look at the different aspects and details of the current project so you can determine where you can use parametric estimation.
In selecting which parts to use the estimating technique, you should ask the following questions.
In determining the scope of parametric estimating, you can also use a work breakdown structure.
Since parametric estimating calculates project estimates using statistical correlation, you need to base your computation on data. A project manager can use data from previous projects found in various sources. Potential data sources include the following.
Now that you have the data, it is time to determine which parameters will correlate with your project's cost or duration requirement. You can test these parameters using statistical analysis or expert judgment. Whether you will use statistical methods depends on the size and complexity of the current project.
After you identify the parameters, you have to test whether they are time or cost drivers in your project using statistical methods and software. This correlation and/or regression analysis is done to select appropriate parameters for the project's estimation model.
After you test the parameters identified, you can now calculate the deterministic or probabilistic estimates for your projects using the parametric estimating formula below.
If your project is not complex, you will first have to calculate your cost or time estimates per parameter unit. There is an additional step before calculating the parametric estimates for large, complex projects where expert judgment will not suffice in estimations.
How do project managers calculate parametric estimates? They can be calculated using the following formula:
E_parametric = a_old/p_old x p_curr
E_parametric is your parametric estimate;
a_old is the historical cost or time;
p_old is the historical value of the parameter, and;
p_curr is the value of the parameter in your current project.
This formula determines the statistical relationship between the historical data and other variables to calculate estimates.
A simple example is a project to create a marketing campaign for a new product. You identify similar marketing campaigns from the past, and break down the size of the team, the length of the campaign, and the office resources that were used. You decide on the cost of each of these variables by breaking them down into smaller units, e.g. cost per person. You then calculate your breakdown per unit, adjust for changes in historical costs, e.g. inflation, and multiply your figures to reflect the increases (or decreases) in the staffing, length of campaign and offices resources needed for the new project.
A project can often be complex and extensive, so regular estimating techniques using judgment won't be enough. In these cases, you should develop a model that estimates costs or duration based on your identified parameters. The model and the results you derive will need back-testing against historical data to ensure the accuracy of the model.
After you develop your model and back-test the results, you can now calculate the parametric estimates by substituting the current project's parameters into the developed model.
As mentioned, parametric estimating is dependent on the size and complexity of the project you are handling. Large projects such as mega-projects will require complex statistical models and even a comprehensive regression analysis for some parameters. These projects may also require project managers to develop algorithms and perform back-testing. Even the most minor mistake in the estimation process can materially impact your project accounting.
On the other hand, smaller projects often apply the rule of three (or probabilistic estimates) when calculating parametric estimates. These projects can also involve expert judgment on determining whether any regression is reasonable or applicable to the current project.
Parametric estimating is a quantitative estimation methodology that is an accurate and exact means of estimating a project's expenses and duration. Some of the benefits of using parametric estimating in project planning are:
Because estimates are derived from past projects, you can reuse your estimates for other similar projects you might handle. You can repeat the process without decreasing the quality of the data. With parametric estimating, you can create a reliable model for reference in your present and/or future projects.
Generating estimates is crucial for a project. Project managers want to ensure that their calculations are not under-estimations, or over-excessive. When you apply parametric estimating in your project, you can be confident that the time and cost values of the project are accurate.
Parametric estimation considers both historical and current data available for use, and their differences. Project managers manually adjust the calculations for the differences between the data, contributing to a clearer model.
Although parametric estimation is applied during project planning, it can also be applied also to various project stages of the project. When applied to different project stages, parametric estimation makes a project coherent throughout its life cycle.
Project planning requires a lot of time and preparation, whether you are looking at a small, large, or even a megaproject - and there's a learning curve when you first start making an initial estimate of costs for a whole project.
You can apply several estimation techniques when estimating your project's budget, timeline, or the resources needed for the project tasks. Other than project estimation techniques, you can also find software that can help you in many aspects of accounting for your project. These tools, software, and techniques are designed to make planning and management of your project more accurate and straightforward.
New to project cost estimation? Check our guides here:
A project dashboard can make or break a project. Learn how great project dashboards look like and what information you should track to avoid overruns.