Last quarter, I had lunch with a guy I'll call Tom.
Tom is the president of a manufacturing business in the Midwest. About $200 million in revenue. Eight hundred employees. Family-owned, third generation. He has been in the role for nine years and at the company for almost twenty. He is, by every measurable standard, a successful operating executive.
We were halfway through our entrees when he pulled out his phone and showed me a Google Doc.
"I started this in 2021," he said.
The doc was titled with a working name and the word Idea in the title. It was a product concept. A real one. He had identified a customer pain point in his industry, sketched out the solution, mapped the market, even priced it. The first entry was dated 32 months ago. The most recent entry was from earlier that morning, on his way to lunch.
"I've added to this thing every couple of weeks for almost three years," he said. "I have not built a single thing."
He looked at me like he expected to be judged. So I want to tell you what I told him, because if you are reading this, there is a non-trivial chance you have a doc just like Tom's. And the thing I told him is the thing you need to hear.
What I told Tom at lunch
The first thing I told him was that he was not the exception. He was the rule.
In ten years of doing this work, I have met dozens of Toms. Different industries. Different revenue bands. Different doc software. Same exact pattern. A real idea. A real market opportunity. A real executive who is, for some reason, unable to move it.
The reason is not what they think it is. It is almost never what they think it is.
The reason is structural.
Every operating executive at a mid-market company is running a system designed for reliability. Predictable revenue. Predictable margins. Predictable people. The whole thing is tuned to deliver the same product next quarter that it delivered last quarter, only slightly better.
That system is excellent. It's also actively hostile to the kind of work a new product launch requires.
A new product needs waste. It needs weirdness. It needs the willingness to throw out a month of work because the customer wants something different. The KPIs are about learning, not output. The people are wired to ask "what if we built this completely differently" rather than "how do we ship the sprint on time."
When you ask the same team to run both, the existing business wins. Every time. Not because anyone is being lazy. Because that's what you hired them to do.
Tom nodded. He had not heard it framed that way.
The four fixes Tom had already tried
He listed them off on his fingers without me asking.
Consulting firm in 2022. Spent low six figures. Got a beautiful deck. The deck is in a drawer. None of his operations team executed on it because they had real jobs.
Development shop in early 2023. Asked him to write a spec. He spent four months writing one. By the time he finished, he realized the spec was wrong, because you can't write a great spec for a product that doesn't exist without customer feedback you don't have because the product doesn't exist.
VP of Innovation hired in late 2023. She was excellent. Also expensive. She needed a team. The team needed budget. The budget needed ROI projections. The ROI projections needed customer evidence. The customer evidence needed a product. There was no product. She left after 18 months.
"Next quarter" approximately 47 times since 2021.
He looked at me. "What am I missing?"
What had changed since he started
Here is the thing I most wanted Tom to understand.
When he started writing in that doc in 2021, the cost of building a real digital product was different than it is today. A team of six. A year. Half a million dollars to find out if it would even work. That is a real bet for a $200 million company. It is reasonable that he hesitated.
But the math has changed. Drastically.
In the last 18 months, the cost of building a credible v1 of most digital products has dropped by roughly 90 percent. The labor is augmentable. The infrastructure is rented. The tools that used to be the bottleneck are now commodities. A product that needed six people and a year now needs one or two operators and a quarter.
Which means the bet Tom rejected three times in the last three years was, today, a different bet entirely. Not a six-figure-and-a-year experiment. A five-figure-and-a-quarter one.
He had been making the right decision for the 2021 economics, and continuing to make it for 2024 and 2026 economics that were no longer the same.
This is not unusual. Most of the executives I talk to are running on outdated math. Nobody warned them the floor had moved.
What we agreed on by the end of lunch
Tom did not need another consultant. He did not need another VP. He did not need another quarter to think about it. He had been thinking about it for 32 months. He needed evidence.
What he needed was a single outside operator who could own the whole thing for 90 days. Strategy, build, launch. Not a deck. A real working version of the idea, in front of real customers, by the end of the quarter. Then look at the data. If the bet was real, scale it. If it wasn't, kill it cheap and move on.
By the end of lunch, that is what we'd agreed to do.
The doc has not been opened since.
The product has.
The thing I want you to walk away with
If you have a doc like Tom's, you are not behind. You are not crazy. You are not the exception.
You are a working executive sitting on a hunch, at a moment when the economics of testing that hunch are the most favorable they have ever been in your career. The only thing standing between you and a real answer is the structure you've been trying to use to get there.
You don't need another quarter. You need 90 days and one operator who can ship the version that proves the bet.
That is the post I wanted Tom to read three years ago, and the one I wanted to write you today.
Same core argument, different doorway: the confessional version · the operating-systems lens.
If that sounds like your situation, book a call. Thirty minutes. No pitch. If we're a fit, we'll talk about the 90-day path to a real v1 in market.