Estimating labour makes or breaks a project

I recently ran into a customer who thanked us for our work on a project. Though he said our labour was wrong, so he had the team double it. He then proceeded to tell me that they lost the $3m project by $300k. Afterwards I wondered, maybe our labour wasn’t so bad?

I get that quite a bit; builders presenting their labour as fact. Whereas the reality is, it is just as much an estimate as ours. So, who is right? The builder’s labour based on experience or our detailed labour build up refined on nearly 4,000 projects?

Labour is people doing tasks. Estimating labour comes down to three steps:

1. Determining the tasks involved
2. Establishing what factors drive the work on the task
3. Measuring those factors

These three steps underpin all labour estimating methods. The main difference is how much you roll up the tasks. The more you combine tasks, the simpler the job appears but the more detail is lost and the broader the assumptions become.

Common labour estimating methods

Based on a project schedule: Estimate the number of weeks for each aspect, multiply by the number of staff on site.

Unit rates: Allow a certain amount of time per building element.

Based on materials: Allow a constant amount of labour against each item in a materials schedule.

Bottom-up task analysis: We call this activity-based costing.

Our experience with ceiling batten labour

When I started YourQS we used hours per m². It was easy. Measure the room area and times by the factor and you have your hours. Then builders started complaining that we weren’t allowing enough time on small rooms.

How we improved

We could have used hours per metre of ceiling batten, but that would still understate small rooms.

I could look at the house and think, “mmm that will take four people five days." But how could I train this to my team?

Instead, we turned to activity-based costing. We analyse the activities and build up the labour times based on these:

1. Get the materials into the room and set up dumpy level
   - x minutes per room
2. Install the perimeter channel/batten
   - x minutes per metre of perimeter
3. Install battens
   - x minutes per m²

Of course, the reality more nuanced than this. There are other factors like materials and build complexity to consider. We represent our knowledge like this:

Use the power of computers

This is very time consuming to do with traditional estimating methods. By harnessing the power of computers and our 3D model, it takes us no time and gives very accurate labour hours that respond with the house design.

A last thought

Track what happened and give feedback to your estimator. It is only by closing the loop can we improve our estimates.