Review Every Job. Improve the Right Part of Your Business.

Every completed job should make your next job better

I've been in the painting business for about 15 years. The first thing I ever learned about painting was how to estimate a job.

My first employer taught me a fairly rigid estimating method based on production rates, square footage, and the amount of work required to complete different tasks.

Fifteen years later, the basic idea is still exactly the same. Figure out how much work needs to be done, estimate how long it should take, calculate what it will cost, and charge enough to make the profit you want. That's it.

I've changed some standards over the years. I've developed different ways to calculate certain types of work. I've experimented with alternative estimating methods, and sometimes I'll calculate the same project two different ways just to compare the results. Usually, they come out pretty close.

The fundamentals haven't changed much because they work. What has changed is my ability to recognize difficult work, estimate complicated setups, anticipate preparation problems, and identify the things that can make a project take longer than expected.

That knowledge comes from experience. And a large part of that experience comes from paying attention to what happens after we sell a job.

Every completed project is an opportunity to improve how we estimate and produce the next one.

But that doesn't mean every completed project should lead to a change in the estimating system. Sometimes everything went exactly as it should. Sometimes the estimate was wrong. Sometimes the painters were inefficient. And sometimes something genuinely unexpected happened. The important thing is understanding the difference.

Estimating is about calculating labor, not guessing a price

When I look at a house, I'm not trying to guess what the customer might be willing to pay. I'm trying to determine how much work the house represents.

I look at the surfaces, preparation requirements, number of coats, access, equipment, and everything else required to complete the job. Then I calculate the labor budget.

My estimating system has documented production standards for different surfaces and tasks, usually expressed in square feet per hour. We also have reference pictures showing examples of what different amounts of labor look like.

An estimator can use those standards to calculate the hours required for each part of a project.

Personally, I often do it differently. I've painted so many houses that I can look at a surface and visualize how long it should take. Sometimes I'll mentally divide a wall into smaller sections. I'll even draw little squares in the air with my finger to help visualize the work. I look at a section and think: that's one hour. That one's two. This one might take three.

Then I add everything together.

I've developed a pretty good sense of how long painting work takes because I've done so much of it myself. And I know I'm a fast painter. So if I think I could complete something in one hour, I might budget two or two and a half hours for it.

For particularly difficult work, I might allow even more.

That isn't an accident. My labor budgets are deliberately generous. A good painter should be able to beat them.

If I budget 100 hours, a competent, productive crew should generally be able to complete the work in something like 50 to 70 hours. That's how the system is supposed to work.

We have productivity bonuses that reward painters for beating their budgets, so the painters have an incentive to work efficiently. When they finish in 60 hours instead of 100, they're happy because they earn a bonus. I'm happy because the company makes more money and frees up production capacity. Everybody wins.

And that distinction is important when we start reviewing completed jobs.

Beating the budget doesn't mean the estimate was wrong

Suppose I estimate a project at 100 labor hours. The painters complete it in 60. The workmanship is good, the customer is happy, and everything was done according to our procedures.

Should I reduce the estimating standard so the next similar job only gets 60 hours? Absolutely not. Why would I?

The system worked exactly as intended. The painters beat their budget, earned their bonuses, and the company made a good profit.

If customers are willing to pay the price we're charging, why would I voluntarily reduce it? I'd be giving up margin to potentially win one more project. And then I'd have another customer to manage, another job to schedule, and more work to produce, possibly for less total profit. That makes no sense to me.

I'd love it if every job could be completed in half its budget. The painters would be ecstatic. And if I suddenly started cutting their budgets because they were performing exceptionally well, they'd have every reason to be upset.

Imagine being rewarded for working efficiently, then discovering that the reward for your good performance is a tighter budget and a smaller bonus. That's not the incentive system I want.

A good estimating budget isn't necessarily the amount of time a good painter will actually need. It's an allowance that lets us price the work profitably while giving productive painters an opportunity to earn more.

That doesn't mean we ignore jobs that finish far ahead of budget. We still ask what went well and whether we can repeat it. If a crew finished in 60 hours, could we organize the next project even better and finish in 55? Maybe. But that's a production improvement, not necessarily a reason to lower the estimate.

When a job goes over budget, find out why

Now suppose we estimate a project at 100 hours and the painters take 120. Something went wrong. But what?

A lot of contractors will immediately assume that the estimate was too low. That's possible. But it's only one explanation.

Maybe one painter was extremely slow while everyone else worked efficiently. Maybe the crew wasted half the morning.

Maybe they didn't have the right equipment. Maybe the foreman organized the work poorly. Maybe the preparation was harder than expected. Or maybe the estimator simply forgot to account for something.

Those are different problems, and they require different solutions. If you don't understand why the job exceeded budget, you don't know what to fix. I separate these situations into three categories.

1. The estimator didn't follow the system

We have an estimating grid, documented production standards, and a checklist of what needs to be inspected. The estimator is expected to use them.

Suppose a house requires 80 hours of painting and preparation, plus another 25 hours for scaffolding, ladder handling, and difficult setups. If the estimator only budgets the 80 hours and forgets the setups, that's not a problem with our production standards. The system already accounts for setups. The estimator simply didn't follow it.

That's a performance problem. We review what happened, retrain the estimator if necessary, and make sure they follow the checklist properly next time.

If the same person repeatedly makes mistakes that cost the company money, there may eventually need to be consequences. That could affect their commission, or in serious cases, their employment. An estimating system is only useful if people actually follow it.

2. The estimating standard was wrong

Now suppose the estimator followed the system perfectly. The grid says a particular task should be budgeted at 10 hours. The estimator enters 10 hours. But the work consistently requires more.

If I inspect the project myself and conclude that I would have budgeted 15 hours, the problem isn't the estimator. It's the standard. We need to correct it.

There's a big difference between a crew taking 12 hours to complete work that should reasonably have been budgeted at 15 and taking 12 hours to complete work that should only require six. The first situation might actually represent good production. The second could be a serious production problem.

You can't tell the difference simply by looking at the variance. You need to understand the work.

3. Production was poorly executed

The third possibility is that the estimate was perfectly reasonable, but the painters didn't execute the job properly.

Maybe they worked inefficiently. Maybe they didn't follow the planned sequence. Maybe the foreman didn't supervise them properly. Maybe they spent far too long moving scaffolding because they weren't handling it efficiently.

If that's what happened, increasing the estimating budget would be the wrong solution. We'd simply be adjusting our prices to accommodate poor performance. Instead, we need to improve production.

Maybe the crew needs better instructions. Maybe the foreman needs to supervise a particular task more closely. Maybe someone needs to demonstrate a better way to perform the work.

Whatever the issue, we correct it where it actually occurred. Fix the estimating system when the estimating system is wrong. Fix production when production is wrong.

The three things that most often cause estimating mistakes

After years of estimating houses, I've noticed that most unexpected labor overruns come from a relatively small number of issues. The first is preparation. The second is access and setup. The third is unexpected material consumption, which can also increase the labor required to apply the coatings.

There are other complications, of course. Certain types of cutting can be exceptionally time-consuming, for example.

But those are the big ones.

Preparation

Two surfaces can look fairly similar and require completely different amounts of preparation. One might sand easily.

Another might have old, failing coatings that require considerably more work. You need to understand what you're looking at.

When I inspect a house, I touch the surfaces. I look at the age and condition of the coating. I often carry a small brush and some sandpaper so I can test the surface. Sometimes I'll spend just a few seconds sanding an area to see how the material behaves. That tells me more than simply looking at it. And it also helps demonstrate to the customer what we're actually going to do.

I've handled thousands of painted surfaces over the years. I can often tell quite a lot just by touching one. That's experience. But the underlying principle is something any experienced painter can apply.

Don't just estimate how much surface there is. Estimate what actually needs to happen to that surface before it's ready for paint.

Access and setup

This was one of the mistakes I made most often when I was learning to estimate. I'd calculate the square footage correctly. I'd calculate the painting hours correctly. Then I'd realize there was an extra day of work just moving ladders, assembling scaffolding, or setting up access.

The painting calculation wasn't wrong. I had simply forgotten part of the job. A complicated setup might require three people instead of two. Some equipment needs to be assembled, moved repeatedly, and dismantled. Sometimes one worker has to assist another rather than paint. Those hours count.

Our estimating system includes allowances for setups, and our estimators have to account for them. If they forget, that's not a reason to change the production rate for painting a wall. It's a reason to make sure they follow the estimating procedure.

Material consumption

Some surfaces absorb far more paint or stain than expected. That can increase material expenses, and sometimes it means additional application work. Other times, the opposite happens. A project goes much faster than expected, or we use far less material than anticipated.

That's part of the business. The goal isn't to predict every unusual situation perfectly. It's to recognize the conditions that tend to create problems and account for them before committing to a price. And when something genuinely unusual happens, learn from it without automatically assuming every future job will behave the same way.

Sometimes the best way to settle an argument is to get on a ladder

Suppose our estimating standard assumes a certain production rate. The painters repeatedly tell me it's unrealistic.

Maybe they're right. Maybe they're not. We can discuss it all day, but eventually there's a pretty simple way to find out.

Go do the work.

Get on a ladder, prepare the surface, and time how long it takes. I might do it myself. Or I'll have someone perform the task while we observe and measure the result.

If I'm an experienced, fast painter and I have to work exceptionally hard just to meet the budget, that's a pretty good indication that the allowance is too tight. If I can comfortably beat it, then we need to understand why the crew is struggling.

Maybe the work is different from what I tested. Maybe they need better equipment. Maybe they aren't following the proper method. Or maybe they're simply not working efficiently.

The point is to replace opinions with evidence. When you disagree about how long something should take, test it. And if the evidence shows the estimating standard is wrong, change it. If the standard is correct, fix whatever is preventing the painters from meeting it.

The estimator's job is to follow the grid

I've trained painters and foremen to estimate jobs. I wouldn't particularly want to train someone who has absolutely no painting experience. There's too much practical knowledge involved.

You need to recognize different surfaces, understand preparation, identify access problems, and know what the work actually involves. Someone who's painted for years already has a foundation. They understand the work. The next step is teaching them how to calculate it consistently.

Our estimating system includes production standards, reference pictures, setup allowances, and a grid where the labor hours are entered. That means the estimator doesn't need to invent a new estimating method. They need to learn how to apply the one we already have.

When I'm training someone, we'll estimate jobs together. I'll demonstrate how I calculate the work. Then I'll have them do it themselves. Sometimes we'll inspect the same house and independently prepare our estimates. Then we compare the results.

Suppose I budget eight hours for one wall and the trainee budgets twelve. We'll go through the calculation. Maybe the production standard is 100 square feet per hour. The wall measures 40 feet by 20 feet. That's 800 square feet, so the budget should be eight hours.

Why did the trainee calculate twelve? Maybe they measured the wall as 60 feet long instead of 40. There's the mistake.

But I don't only review the calculations where our answers differ. I review the ones where we agree, too. Maybe we both calculated eight hours, but the trainee used the wrong production standard and made another mistake that happened to cancel it out.

Getting the right answer by accident isn't good enough. I want to know that they followed the correct process. So we go through everything. I demonstrate. They perform the calculation. I review their work and explain any corrections. Then they do it again. We repeat that until they can estimate properly.

And I'm willing to be challenged. Sometimes they might notice something I missed. I make mistakes, too. If someone disagrees with me, I want them to explain why. I'll listen. But they need to have a reason.

Once they're trained, the expectation is clear. The estimator follows the system. They don't invent their own production standards. If the system produces incorrect budgets, we correct the system. If the estimator isn't following the system, we correct the estimator.

Don't change the system every time something unusual happens

Sometimes a project simply goes badly. The preparation turns out to be unusually difficult. A surface absorbs far more paint than expected. A particular section requires an awkward setup that wasn't obvious during the estimate.

Those things happen. And sometimes the opposite happens. A project goes much faster than expected, or we use far less material than anticipated.

One unusual result doesn't necessarily mean the estimating system is wrong. If we've completed dozens of similar jobs using the same standards, and almost all of them performed as expected, I'm not going to rewrite the system because one project was unusual.

But if the same problem keeps happening, that's different. If a particular type of preparation repeatedly takes longer than our standard allows, we need to investigate. If the evidence shows the allowance is too tight, we change it.

If a certain type of setup consistently requires more people or more time, we account for that. If our material assumptions are repeatedly wrong, we reassess them.

The objective is to make the system more reliable over time. Not to chase every unusual result. And there's another reason to stick with familiar types of work.

After 15 years, there aren't many painting surfaces or situations I haven't encountered before. I know what normal looks like. That makes it much easier to recognize when something is genuinely unusual.

It's also why I prefer selling projects we understand. The more predictable the work, the more confidently we can estimate it and protect our margins.

Every completed job gets reviewed

We don't wait until the end of the summer to figure out what went wrong. That would be ridiculous. If we're repeatedly making the same estimating mistake in May, I want it corrected in May. Not in October, after we've sold another twenty jobs using the same bad assumptions.

Every Monday, we have a management meeting. The owner, project manager, foreman, and sales representative attend. We review the jobs from the previous week. Usually, each job takes no more than a couple of minutes.

Most are uneventful. The work went according to plan. The customer is happy. The crew performed well. There aren't any major issues. Perfect. That's how it should be.

But we still ask how the job went and whether there's anything worth learning. Maybe the crew handled a difficult setup particularly well. Maybe they found a more efficient way to sequence the work. Maybe something went wrong, but they still finished comfortably under budget.

We want to understand those things. And if a problem occurred, we decide what needs to change.

Suppose a crew wasted half a day moving scaffolding. Did they know how to move it efficiently? Were they properly supervised? Was the equipment appropriate? If the problem was poor execution, the foreman makes sure it doesn't happen again. We don't need to change the estimating grid.

But if we discover that the setup genuinely requires more labor than our system allows, that's an estimating issue. We correct the standard. Then we communicate the change to the people involved and update the documentation.

That's the process. We identify the problem, agree on the correction, implement it, and move on. The detailed management-meeting system is a subject of its own, but the important point here is that the review happens every week, while the information is still fresh and the lessons can be applied immediately.

Even good jobs can teach you something

A common mistake is only investigating projects that lose money. I don't see it that way. We review every job.

Suppose a crew completed 100 budgeted hours in 70. That's good. But could we have done it in 67? Could we have done it in 60? Maybe there's a better way to organize the work. Maybe a particular sequence saved time. Maybe the foreman did an excellent job preparing the site before the painters arrived.

If something worked exceptionally well, I want to understand it. Because I'd like every project to go that way.

And suppose something went wrong during that same job. Maybe the painters wasted time on a setup but still finished well under budget. The good financial result doesn't mean we should ignore the mistake. We should still fix it.

Otherwise, the same mistake might happen on another job where we don't have as much room in the budget.

The objective isn't merely to identify unprofitable projects. It's to improve the operation continuously. That's how I approach almost everything in business.

You do something. You observe the result. You identify what worked and what didn't. Then you improve the process.

It's no different from a hockey team reviewing the previous night's game. Even if they won, they still watch the video.

They want to understand what they did well and what they could have done better. A painting company should operate the same way.

What happens when we discover an estimating error?

Let's say we discover that our standard for a particular task is too optimistic. We investigate the work. We discuss it with the estimator and foreman. If necessary, we test the production rate ourselves. Then I decide whether the standard needs to change.

In my company, only the owner has the authority to approve changes that affect how much money comes in or goes out. An estimator can't independently decide to change the company's pricing assumptions. Neither can a foreman.

They can bring me information, explain the problem, and recommend a correction. But the decision is mine.

Once we've agreed on the change, we update the estimating documentation and make sure everyone involved knows about it. That's it.

There's no need for an elaborate annual estimating audit. We already review our jobs every week. We already know when projects go over budget. We already discuss recurring problems. If the system needs correcting, we correct it now.

Of course, there's one frustrating part. What about the quotes we've already sent using the old standard? Sometimes we can try to revise an outstanding quote, but that generally doesn't go particularly well.

And if the customer has already accepted the price, we've made a commitment. Usually, we have to live with it. That's the cost of the mistake. We can make sure it doesn't happen again, but we can't always recover the money on work we've already sold.

Which is another reason to correct problems as soon as we discover them.

If you can't estimate the work confidently, don't pretend you can

One of the easiest ways to lose money is to accept a project you don't properly understand and give it an optimistic fixed price.

Sometimes you inspect a house and encounter a section that's unusually difficult. Maybe the preparation is unpredictable. Maybe the access is complicated. Maybe you can't confidently determine how much work will be required.

You have two reasonable options. The first is to increase the labor allowance. If there's significant uncertainty, I might double or even triple the hours I'd normally allow for that section.

I'd rather quote a higher price and lose the job than sell it too cheaply and lose money producing it.

The second option is to charge by the hour. If I genuinely cannot predict how much work something will require, hourly billing might be the only sensible way to handle it.

Sometimes a customer will ask how many hours I expect it to take. Well, if I knew that with enough confidence, I'd give them a fixed price. That's the whole problem. I don't know. And I'm not going to pretend I do just to win the project.

You can make good money painting houses without accepting every uncertain job that comes your way. In fact, one of the reasons our estimating system works is that we concentrate on work we know how to produce.

We understand the surfaces. We understand the preparation. We understand the setups. And we have years of experience telling us what reasonable labor budgets look like.

When something falls outside that experience, I price the uncertainty accordingly.

The purpose of the system is to make the next job better

Over the years, I've made plenty of estimating mistakes. I've underestimated preparation. I've forgotten about complicated setups. I've misjudged how much material a surface would consume. I've had jobs take much longer than expected. I've also had jobs go exceptionally well. That's business.

The important thing is what happens afterward. If the estimate was wrong, I want to know why. If the estimating standard was wrong, I correct it. If the estimator didn't follow the system, I address that. If the painters performed poorly, I correct production. If something went exceptionally well, I want to understand how we can repeat it.

And if the job was simply unusual, I take note of it without unnecessarily changing a system that already works.

There's no need to make this complicated. We have an estimating system. We have production budgets. We track actual results. We review completed jobs every week. And when something needs correcting, we correct it.

That's how a business improves. Not by sitting down at the end of the season and discovering that you've been making the same mistake for six months. Not by endlessly changing your prices every time a crew finishes early or late. And not by blaming the estimator whenever production exceeds budget.

You improve by understanding what actually happened, correcting the right problem, and making sure the next job benefits from what you learned.

After 15 years, the fundamentals of my estimating method are still largely the same as when I first learned them. The difference is that I've had thousands of opportunities to understand the work better. And I intend to keep learning from every job we complete.

Because even when the system works well, there's always room to make the next job a little better.

These are results and experiences from my own company, not promises of what will happen in yours. AI helped organize the writing; the operating experience, reasoning, and final responsibility are mine.