The secret weapon to precise project planning: Parametric estimating (with examples!)

What is parametric estimation?

Parametric estimating is a form of project estimation that works for time, resource and cost estimates.

It uses parameters (characteristics) to create estimates based on what you already know, so in that respect it’s useful for a project manager because it’s reliable. You get accurate data because they are data-based and thought-through – it’s a quantitative approach.

Oh, and it’s one of several project estimation techniques covered by the Project Management Professional (PMP)® exam too, so if you are studying, you’ll need to understand the concepts and how to apply it so you can be confident on your test.

Definition

Estimating approach that uses a statistical relationship between historical data and other variables to calculate an estimate.

Dictionary of Project Management Terms, 3^rd Edition, J. LeRoy Ward

When to use it

Parametric models are used in three situations:

1. When you have data to use for modelling (get that from historical information from similar projects)
2. The model is scalable (in other words, the parameters remain the same regardless of how many units of work are required)
3. It’s easy to work out the parameters required: the work is easily quantifiable into recognizable units e.g. hours, dollars, meters etc.

It’s good for repetitive tasks, but if those conditions aren’t met, you are going to struggle to use this technique.

For that reason, it’s not a method I tend to use very often, especially on complex projects. Mostly, I’m estimating knowledge work or tasks where analogous estimating (where you look at past projects and see how long similar tasks took) is a more reliable technique.

However, if your new project has tasks which have a statistical relationship: number of square meters ploughed in 30 minutes, square footage prepared in construction, number of lines of code written in an hour by an experienced developer, then use it to get a better estimate than simply ‘expert judgment’ which for most people means guessing.

Parametric estimating in the PMBOK® Guide

This statistical method gets a mention in the PMBOK® Guide as a way to estimate activity durations, costs and resource effort.

The PMBOK® Guide talks about it as a technique to use either for a total project, or for working out the estimates related to parts of a project from the work breakdown structure. You can combine it with other techniques to calculate the overall project cost or duration.

Parametric estimating formula

In simple terms, multiple the quantity of the work to be performed by the number of labor hours per unit of work (or cost, or whatever your parameter values are). Then you get an accurate estimate.

It’s clearer to see how this works in practice with a couple of examples, so that’s what we’ll see below.

How to do it

First, establish the parameters. Are you looking for time estimates or cost estimates? Make an informed decision for this project stage e.g. units of work required.

Next, look at previous projects: it’s always helpful to see what happened in the past and what you can learn from the past.

Work out how many units of work are required.

Multiply units of work required by cost/time/effort for each unit. Voila! There you have it: resource requirements as validated by the estimation process.

Using any new project management technique is a learning curve when you get started, so don’t worry if it seems tricky at first – it’s actually very easy once you get used to it.

Be sure to capture what you did and keep that information for your next project: even if there are unique complexities in this particular task, it will be helpful to look back on your past experience.

Parametric estimating examples

When you’ve got the data to use at task level, the parametric method gives you a high accuracy of estimates because they are grounded in data you already know to be true.

Here are some examples to show you how it works.

Example 1

Here’s an example:

If it takes one person one hour plant 20 potatoes, it will take two people two hours to plant 80 potatoes.

Even if this turns out to only be a rough estimate based on their skill level (a novice potato planter will take longer per potato than an expert planter, for example), it still gives you a very good starting point.

Note: some things can’t be estimated this way. You know the old saying: it takes one woman 9 months to have a baby. You can’t add another woman and get the baby in 4.5 months.

Example 2

Here’s another example for working out time requirements:

A network engineer is capable of laying 25 meters of fiber optic cable per hour, on average. The project requires her to lay 2,000 meters of cable.

2,000 / 25 = 80 hours of work

That doesn’t include any travel time to the site, or breaks. As the project manager, I would look to see if we could assign a couple of network engineers to get the cable in faster.

Your models can build in all kinds of variables for improved accuracy, but at the most basic levels, those examples show how to work out project estimates.

Example 3

Here’s another example, this time for a cost estimation:

It costs £500 for a day of consultancy. Your project is being outsourced to the consultancy and they have given you a Gantt chart saying the work will take 10 days, spread over 7 weeks.

The formula for cost estimates would be 500 x 10 so the total cost of the project is £5,000.

Then you’d track the actual cost and adjust as you go based on performance and what you learn. Rework your estimate at completion during the project so your financial tracking stays accurate.

Calculating total project cost

It’s unlikely that you can estimate your whole project with parametric calculations. It’s appropriate to use for certain tasks where you have the data and this information can then be aggregated.

However, you’ve probably got tasks in your project where this way of estimating simply isn’t accurate or advised.

The size of the current project most likely determines how many and what type of techniques you are going to use to get variable data and precise estimates. Be prepared to mix and match.

Read next: Project Resource Management: The Ultimate Guide on How to Master It.

Limitations of parametric estimating

As with all project management techniques, this is one to use under advisement. The basic formula of ‘time it takes to do the work’ x ‘amount of work required’ might seem straightforward but it might not translate to the real world.

For example:

• Those potato planters might get exhausted and need a break after every 20 minutes of digging
• One of them might plant faster than the other due to having more experience or better tools
• The network engineer might be working in a building of historical interest and need to use different techniques for some of the cable-laying to preserve the look of the building
• The projects might have variable cost: perhaps laying cable in the summer costs more than in the winter and you have to factor that in because the historical building is occupied and the disruption is greater
• If you have little information (for example, you don’t have the data points for how long it takes to plant a potato), you can’t use this technique until you have them
• You need reliable data: if you’ve picked the consultancy rate of a blog post on the internet and not the actual proposal from your consultancy firm, you could be in for a shock when the bill comes.

The accuracy of parametric estimates is only going to be as good as the project parameters and statistical models created by the project team. Take care in getting the basics right and you will be rewarded with higher levels of accuracy.

However, in the absence of nothing better, parametric estimating gives you an overall, high level estimate and still allows you to add in contingency or other variables to get a more accurate result. It’s definitely one of the estimating techniques to keep in your toolkit for when your project needs it.

Read next: Project Estimating: The only guide you need