Murphy Door runs three American plants. Ogden, Utah, our headquarters, where we do finishing and door assembly. Lexington, Kentucky, our East Coast door plant and fulfillment operation. Rigby, Idaho, our hardwood mill, where we cut material, floating shelves, stair risers, veneered products, mantels, cabinet doors, select door lines, custom hardwood supply, and a furniture line that now includes nightstands, dressers, armoires and vanities. Murphy Door Machine builds the hardware in house. We have shipped over two hundred thousand orders. Every one of them was made in America.
I put those numbers up front because most conversations about American manufacturing skip past the numbers and go straight to opinions. Opinions are cheap. Plants are expensive. If you are building a company that actually makes things, the opinion part matters much less than most people writing about manufacturing want to believe.
This is the honest version of what it takes to build American manufacturing that works. This is what I actually run. Call it a plant walk instead of a policy essay.
You can't run a plant from a spreadsheet
The first thing that surprises founders who move into manufacturing is that a plant is a physical place, and physical places don't respond to spreadsheets. You can plan a plant on paper for a year. The plant will do what the floor does, not what the plan says.
What that means in practice is that a plant manager is worth more than most founders realize, and the founder needs to be in the plant enough to know why. I spend real time in Ogden. Not on tours. In the plant. Watching the line, listening to the operators, asking questions I have already asked and getting new answers because the answer is different this week than it was last week.
A plant that works is a plant where the operators know things the plant manager does not know, and the plant manager knows things the founder does not know, and all three of them have a way to move information up and down without anybody losing face. If any of the three layers goes silent, the plant starts making bad parts within a month.
Three plants isn't three times the work
Three plants run closer to seven or eight times the work of one plant, because now every decision has to be evaluated against three sets of local conditions. Ogden runs different than Lexington. Lexington runs different than Rigby. If you try to standardize everything, you lose the local advantages of each plant. If you try to let each plant do whatever it wants, you lose the ability to move product between them when demand shifts.
The answer is a small set of things that must be the same across all three plants, and a wide set of things that can differ. Quality specs must be the same. Safety standards must be the same. How the plant celebrates a birthday, what the break room stocks, when the shift meeting happens, those can be local. The mistake most multi-plant companies make is standardizing culture and letting quality drift. It should be the opposite.
The hardest thing about American manufacturing is hiring
Everybody who has never run a plant thinks the hardest thing about American manufacturing is cost. Cost is a real problem, but it isn't the hardest one. The hardest problem is hiring people who want to build things with their hands and who show up on time for the shift.
Two full generations were told that a factory job was the wrong thing to want, and now those generations are running the country's talent pipeline. Every plant in America is competing for the same shrinking pool of trades workers, and the plants that win are the ones that treat the shift work like real work, not like a fallback.
We pay well. We schedule fairly. We invest in tools that make the shift less punishing. We celebrate the people who become the best at what they do. We don't talk down to the floor from the office. When somebody on the line has an idea about a fixture that would speed up the finishing bay, we build the fixture the same month, and we tell everybody who thought of it.
None of that is a program. It amounts to a set of daily choices that add up. When I walk through Ogden, most of the operators know me by name and I know them. Not because I am unusually good at names. Because I go in there enough that we have actually talked.
Why we build the hardware in house
Murphy Door Machine builds our hardware in house. The obvious question is why. The received wisdom says hardware is a commodity, hardware is cheaper overseas, hardware isn't the story. In our shop hardware is exactly the story, and here is why.
A hidden door is only as good as the hinge and the pivot. When we sourced hardware overseas early on, we found that our failure mode was almost always the hardware, not the door. A door made of American hardwood, finished in Ogden, hung on a cheap import hinge, still carries our name, but it carries a defect that will show up in year two and cost us the customer.
Building the hardware in house means we control the tolerances, we control the metal, and we control the finish. It also means we can iterate on the hinge design without waiting six months for a container ship. When a customer service call comes in with a complaint we have never heard before, we can be in the tool room the same week with a fix on the machine, not on a design that ships next year.
American manufacturing works when you refuse to outsource the parts of the product that determine whether the product actually works. Everything else can be reconsidered. The parts that fail first should be built by you.
The numbers that matter
Founders who look at manufacturing from the outside always want to know margin. Margin is a fine number. It is not the most important one. The most important numbers in a plant are on-time delivery, first-pass yield, and customer complaint rate by unit shipped.
On-time delivery tells you whether your plant is doing what it said it would do this week. If on-time drops, everything downstream drops with it. Sales gets embarrassed. Customer service gets flooded. The finance team ends up carrying receivables longer than they planned. On-time delivery is the first number I look at every morning in every plant.
First-pass yield tells you what percentage of doors leave the line the first time without needing rework. Rework is where margin dies. Any door that goes back through the finishing bay costs us the labor twice and the material once, and the door that took its place in the queue is now late. First-pass yield is the number that tells you whether the operators have the tools and training they need.
Customer complaint rate per unit shipped is the number nobody outside the company sees, but it is the truest signal we have. Complaints go up before margin goes down. Complaints go up before reviews go bad. Complaints go up before returns spike. If we watch complaints weekly, we catch the problem three months before it hits the P&L, which means we can fix it in the plant before it becomes a spreadsheet problem.
The recognition that actually helps
Murphy Door has been on the Inc. 5000 seven times, listed in Fortune's 2025 Most Innovative Companies, and won Best of IBS in 2022 for the Flush Mount Pantry Door from the National Association of Home Builders. I was named an EY 2024 Mountain West Entrepreneur Of The Year Finalist for the work at Murphy Door. Those are real recognitions and they mean something. They also don't build a single door.
What awards do is give a plant a reason to celebrate together, and give hiring a reason to move faster. When we hang a Best of IBS certificate in the finishing bay, the people who did the work know they did the work, and the next person we hire knows what kind of place they are coming to. That is what recognition is for. That flag is a plant morale tool. It is not there for marketing.
Utah Business named me CEO of the Year in 2026. That belongs to the plant, not to me. The plant is what won it. If Ogden had not been running well, no award would have shown up. Same in Lexington, same in Rigby. The recognition follows the work. The work does not follow the recognition.
What I tell founders who want to build something
If you want to run an American manufacturing company, don't start with a plant. Start with a product that people want, and build the plant around the product. Most first-time manufacturing founders build the plant first, discover that the product is not right, and then have a plant they can't repurpose. That is the most expensive mistake I see, and it happens every year to founders who read too many articles about how America needs to build again and not enough articles about how America builds when it is done well.
Start with orders you can fulfill from a smaller facility than you think you need. Prove the product with customers who paid you real money. Then, only then, start signing leases for square footage. Every square foot of plant you sign for before you have proven the product is square footage that will haunt you for years.
American manufacturing still works when you build it right. The three plants prove it every week. Two hundred thousand orders shipped is a count, not a projection. If you want to build in America, build. But build the product first and the plant second. That is the order that actually works.