I have spent most of my career in industries that do not show up in TechCrunch. Sign manufacturing. Commercial real estate. Self-storage. Equipment rental. Direct lending.
None of these are glamorous. All of them are operationally complex. And every single one of them is being reshaped by AI right now, not by the technology companies building the models, but by the operators running the businesses.
That distinction matters more than most people realize.
The Misconception About Who Benefits from AI
The default narrative is that AI is a technology story. That the winners will be the companies with the biggest models, the most compute, the deepest engineering teams. And for the model layer, that is true. OpenAI, Anthropic, Google, and the rest are competing on infrastructure, and that race requires billions of dollars.
But the application layer is a completely different game. The companies that will extract the most value from AI are not the ones building it. They are the ones that know their own operations well enough to rebuild them.
That is an operator skill, not a technology skill.
I grew up in a sign shop. I learned purchasing, manufacturing, sales, and operations by doing the work. When I eventually bought the company and scaled it from $6 million to $55 million in revenue, the edge was never a proprietary technology. It was knowing the business cold: where the waste was, where the margin leaked, where one process change could unlock a step function in throughput.
AI does not change what makes a good operator. It changes what a good operator can do.
Why Operators Have a Structural Advantage
There are three reasons why traditional business operators are better positioned for AI than most people think.
1. They know where the waste is
The hardest part of deploying AI is not choosing a model or writing a prompt. It is knowing which workflow to point it at. That requires deep domain knowledge. An AI consultant can walk through your business and identify obvious candidates. An operator who has lived in the business for a decade can identify the ones that actually matter.
The difference between "this task is repetitive" and "this task is the bottleneck that limits how fast we can grow" is the difference between a demo and a result.
2. They own the decision rights
In a large enterprise, deploying AI against a core workflow requires approval from legal, compliance, IT, and at least two layers of management. That process takes months. Sometimes it takes years.
An operator in a mid-market business can decide on Monday and deploy on Tuesday. That speed advantage is enormous. AI is moving fast enough that the company that implements this quarter has a meaningful edge over the company that is still evaluating vendors.
I rebuilt the accounting workflow inside one of my portfolio companies in a week. Not because the technology was complicated, but because I had the authority to change the process, I understood the process, and I did not need anyone's permission to try.
3. They can iterate without a committee
The first version of any AI workflow is never the final version. It takes three or four iterations to get it right. Each iteration requires someone who understands the business well enough to evaluate the output and adjust.
That feedback loop is fast when the person running it is the operator. It is slow when it has to go through a project manager, a vendor, and a steering committee. The operator's iteration speed is the real competitive advantage.
The Mid-Market Is the Most Underleveraged Segment
Enterprise companies have AI budgets. Startups are built on AI from day one. The mid-market, companies doing $5 million to $100 million in revenue, is stuck in the middle.
These businesses are too large for the free tier of consumer AI tools and too small for enterprise AI deployments. They do not have data science teams. They do not have Chief AI Officers. Most of them do not have a formal technology strategy at all.
But they have something more valuable: operators who know every workflow in the business by name, who have the authority to change those workflows, and who are motivated by the economics of doing more with less.
The mid-market operator who treats AI as an operating layer, not a line item, will compound faster than the enterprise company that treats it as a digital transformation initiative.
What "AI as an Operating Layer" Actually Means
I use this phrase a lot, so let me be specific about what it means.
Most businesses today use AI the way they used spreadsheets in 1995. They have one tool, they use it for one task, and they think they are done. That is AI as a point solution.
AI as an operating layer means it is embedded in how the business runs. Not just one workflow. The operating rhythm.
Inside my own portfolio, AI touches:
- Accounting: automated categorization, anomaly flagging, and month-end reconciliation
- Property management: lease abstraction, tenant communication, maintenance dispatch
- Storage operations: pricing optimization, occupancy forecasting, lead follow-up
- Equipment rental: dispatch coordination, utilization tracking, preventive maintenance scheduling
- Lending: deal screening, document extraction, risk summary generation
No single one of these is revolutionary. But together, they change the cost structure of the business. Work that used to require three people and five steps gets done by one person and two steps. That delta compounds every month.
The operators who build this layer first will be structurally cheaper, faster, and more responsive than their competitors. And that advantage widens over time, because each workflow you automate frees up capacity to automate the next one.
You Do Not Need to Be Technical
This is the part that surprises people.
I am not a software engineer. I have never written production code. But I have built financial models, pricing engines, operational dashboards, and internal tools using AI coding tools, directing the work rather than writing it myself.
The skill is not coding. The skill is knowing what you want the system to do, being able to describe it clearly, and being willing to iterate until it works. Those are operator skills. They are the same skills you use when you are training a new employee, scoping a construction project, or negotiating a lease.
If you can describe a process in plain language, you can build an AI workflow around it. The tools have reached the point where the bottleneck is clarity of thought, not technical ability.
The Window Is Open, but It Will Not Stay Open
Right now, most mid-market businesses have not started. That is the opportunity. The operator who builds AI into their operations this year will have a two- to three-year head start on the ones who wait.
Two to three years does not sound like much. But in a business doing $10 million to $50 million in revenue, the compounding effect of faster operations, lower overhead, and better decision-making across every workflow is the difference between growing 8 percent a year and growing 25 percent.
The models will keep getting better. The tools will keep getting easier. But the advantage goes to the people who start now, because the real learning is not in the technology. It is in understanding how AI fits into your specific business, your specific workflows, your specific team. That understanding only comes from doing the work.
The operators who treat AI as their job, not someone else's job, are the ones who will win.