AI and the Remodeling Industry
Your Business Model Is a Math Problem. Solve It Before You Automate It.
Ask a contractor what their business model is and you’ll usually hear something like “we do kitchens and baths” or “we’re high-end remodeling.”
That’s not a model. That’s a description.
A business model is a math problem. It tells you exactly how many leads, appointments, sales and jobs it takes to produce the profit you want. If you can’t write it on one page, you don’t have one.
And if you don’t have one, AI has nothing to improve. This is part of a series on why AI is not your business model.
Start at the bottom line and work up
Most owners build their plan top-down. “We want to grow revenue next year.” Then they hope the profit shows up.
Flip it. Start with net profit.
What do you want to keep? Not revenue. Profit. The money left after every job is paid for and every bill in the office is covered.
From there, work backward:
- Net profit target. What you want to keep.
- Overhead. Everything it costs to keep the doors open: office, salaries, insurance, vehicles, software, marketing.
- Gross profit needed. Net profit plus overhead. That’s how much margin your jobs have to produce.
- Gross margin. The percentage you actually keep on each job after materials, labor and subs.
- Revenue needed. Gross profit divided by gross margin.
- Jobs needed. Revenue divided by average job size.
- Sales needed. Usually the same as jobs, minus cancellations.
- Appointments needed. Sales divided by close rate.
- Leads needed. Appointments divided by set rate.
- Marketing budget. Leads times cost per lead.
Now you have a model. Every number connects to every other number. Change one and you can see what happens to profit.
Stress test it
Once it’s on paper, you can stress test it. And this is where it gets interesting.
What if close rate drops? You need more appointments to sell the same number of jobs. More appointments means more leads. More leads means more marketing spend just to stand still.
What if gross margin slips? You need more revenue to produce the same gross profit. That means more jobs, more appointments and more leads, all to make the same money you were already supposed to make.
What if overhead creeps up? Every dollar of new overhead comes straight off the bottom line unless every other number gets better.
This is why I say the net profit doesn’t lie. The model shows you exactly which lever matters most, and how badly a small slip hurts.
Know which business you’re in
Remodeling is not new construction. Home improvement is not remodeling. A one-day bath company runs a completely different model than a design-build firm doing whole-home renovations.
Different job sizes. Different cycle times. Different margins. Different lead costs. Different overhead structures.
If you’re comparing your numbers to a company in a different model, you’re going to make bad decisions. Know which one you are, and compare yourself to companies running the same model.
Then point AI at the weak number
Once the model is on paper, AI finally has a job.
Weak close rate? Use AI to sharpen sales training, build better pre-appointment briefs and tighten follow-up.
Margin slipping? Use AI to check estimates against your target margin before they go out and to catch change orders that never got written.
Overhead creeping? Use AI to take repetitive admin work off the office so you stop hiring for volume.
Cost per lead too high? Use AI to respond faster so you stop wasting the leads you already paid for.
Every one of those is a direct line from an AI project to a number in the model. That’s the only kind of AI project worth building.
Write it this week
One page. Ten numbers. Work backward from profit.
If you can’t fill in a number, that’s your first project. Not an AI tool. A spreadsheet and an honest conversation with your bookkeeper.
The model is the business. Everything else, including AI, is how you run it better.
Next in the series: the five disciplines that execute the model.
