How to build a labour estimate you can actually track against

Key Takeaways

• You don't need a dedicated estimator to build a usable labour estimate, just a number that's consistent from job to job.
• Breaking work into areas, cost codes or elements gives your estimate the same structure your timesheets need to be measured against later.
• Historical job data, even a rough memory of past jobs or an old spreadsheet, is enough to start from. It doesn't need to be perfect to start.
• Every completed job should feed back into the next estimate, so accuracy compounds over time instead of starting from zero on every quote.

Most subcontractors without a dedicated estimator already quote jobs confidently, based on years of experience and a good instinct for what a job should cost. That's real knowledge.  

However, quoting a price and building a labour estimate aren't the same thing. A quote can be accurate without ever being broken down into the hours and structure behind it, which means there's often nothing written down to check a job against once it's underway.

You don't need a dedicated estimator to build a labour estimate worth tracking against. What you need is a number that exists before the job starts and gets built the same way every time. Most businesses already have everything required to do this, it's usually sitting in old job files, invoices or simply in the memory of whoever quotes the work.  

What a labour estimate needs to include

A labour estimate is a number, or a set of numbers, stating how many hours a job should take, written down before the job starts. It doesn't predict the future perfectly and it doesn't need to. It simply needs to exist, and it needs to be specific enough that once the job's underway, you can compare actual labour hours against what you expected.

It also needs to answer more than one question. A labour estimate that only says "the whole job should take 400 hours" is technically still an estimate, but it hides where those hours are meant to go. A useful one breaks that figure down, so that when a job runs over, there's a way to see which part of it ran over, not just that it did.

How that number gets used once it exists, comparing actuals against the estimate, drilling into areas, sub-areas, cost code sand elements is a separate activity. But before you can extract the value from your productivity number, you must build out your estimate in the first place.

Breaking the job down into areas, cost codes or elements

The single biggest difference between an estimate that's useful and one that isn't is whether it's broken down into more granular sections. A total figure for an entire job covering all hours from start to finish is better than nothing, but it hides everything you'd actually want to know. If the job runs over, a single total tells you it went over but it doesn't tell you where.

Breaking the estimate into smaller components doesn't require a formal system. Putting it simply, it requires you to split your estimate along whatever lines your business already thinks in.  

• Area and sub-area, e.g. ground floor, level one, if the job's a multi-storey build.  
• Cost code, e.g. formwork stripping or formwork erection, if your trade already prices work that way.  
• Element, e.g. a lift core or a slab, if your business tracks at that level of detail

You don't need all three layers. Some businesses estimate by area and cost code together. Others use cost code alone and get useful insight from that on its own. What matters is picking a breakdown that matches how your business already talks about a job.  

Turning historical job knowledge into a number

Most subcontractors already have the raw material for an estimate. It's probably sitting in old invoices, completed job files, a spreadsheet buried in a shared drive somewhere, or simply what the director remembers about how long the last three similar jobs took. It’s sitting in the wrong place but it’s still a legitimate starting point.

The estimation process is straightforward. Look at a handful of past jobs similar to the one you're about to price and work out roughly how many hours per unit of work they took, e.g. cubic metres of concrete per hour, square metres of formwork per day, tonnes of steel per shift or whatever unit matches your trade.  

Take the average of all your jobs – that’s the starting estimate for your next one. Then, adjust for anything unique or different, e.g. harder access points, more complex pours, tighter programs, etc.

This won't be as precise, but the point isn't precision on day one. It's having a number that exists, that's grounded in what your business has delivered before, rather than a guess made fresh for every quote with no memory of the last one.

This same process works at the more granular level too. Rather than averaging hours across a whole job, pull the past data for a specific area, cost code or element instead, formwork erection specifically, or just the ground floor pours, if that's the breakdown you settled on earlier. The more your historical data is already broken down this way, the more granular your new estimate can be from day one. If your past jobs were only ever recorded as a single total, start there, and let the breakdown get finer as you build up job history recorded at that level.

How good does a labour estimate need to be?

There's a version of this that goes wrong – an estimate built once, based on a rough recollection of past jobs, then never revisited or refined. An estimate that is never revisited isn’t as relevant or useful as one that gets sharpened with every completed project.

What disqualifies an estimate is its inconsistency, built differently every time depending on who's quoting and influenced by how busy the business happens to be that week.

A business that estimates the same way every time, using the same breakdown and the same method for pulling in historical data, is in a far stronger position than a business that never documents anything at all. Consistency is what makes the data usable, precision can come later.  

Why every job should feed the next estimate

Every completed job sharpens the next estimate. Actual hours compared against what was expected are a far better input than memory or gut feel.  

Once a job's finished, you have real data, actual hours against actual quantities, broken down the same way your estimate was. That's a far better input for the next quote than memory or gut feel. Over time, this becomes a genuine asset. A business that's tracked three formwork jobs on similar structures knows its real productivity rate for that kind of work. That number only gets better every time it's used.  

Once the groundwork is in place, you're building a personalised dataset that belongs to your business, not an industry average or a guess.

Getting started this week

If your business doesn't have an estimating process today, don't try to build something exhaustive before the next job. Pick the job coming up, break it into whatever areas or cost codes your business already thinks in, pull together a rough hours-per-unit figure from two or three recent comparable jobs, and write that number down somewhere before the job starts. Refine it from there, job by job – don’t wait until it's perfect to begin.

‍

Construction Labour Estimate FAQs

How do I build a labour estimate without a dedicated estimator?
Building a labour estimate without a dedicated estimator starts with breaking the job into whatever areas, cost codes or elements your business already thinks in, then pulling a rough hours-per-unit figure from two or three recent similar jobs. A director or operations manager can build this informally, provided it's done the same way every time, so the result stays comparable from job to job.

What information do I need to build a labour estimate?

Building a labour estimate mainly requires historical data from past jobs, hours worked against quantities completed, broken down by whatever areas, cost codes or elements the current job will use. Old invoices, completed job files or a director's recollection of recent similar jobs, are enough to start from, since the number gets sharper with every job it's compared against.

Does a labour estimate need to be broken down by area or cost code, or is a total figure enough?

A total labour hours figure for the whole job is better than no estimate at all, but it hides where a job runs over if the number turns out to be wrong. Breaking a labour estimate down by area, cost code or element, even just one of these layers, shows specifically where a job diverged from what was expected, rather than only that it diverged.

How accurate does a labour estimate need to be to be useful?

A labour estimate doesn't need to be highly precise to be useful, it needs to be consistent. An estimate built the same way every time, using the same breakdown and the same historical data, gives a business something reliable to measure against, even if the underlying number is rough. What undermines a labour estimate is inconsistency, not imprecision.

How does a labour estimate improve over time?

A labour estimate improves each time a job is completed, and its actual hours are compared back against what was estimated, since that comparison sharpens the hours-per-unit figure used for the next similar job. A business that does this consistently across several jobs builds its own real productivity data, rather than relying on memory or gut feel for every new quote.