Twenty million in ARR. Five employees, plus a fleet of managed agents. Do the division: four million dollars of revenue per person.
That is the 2-year target we are building PerformanceLabs against. Will we get there? Fingers crossed. What matters is that today it’s possible. Two years ago, the same claim would have made me sound like an out-of-touch founder.
SaaS Capital surveyed more than 1,000 private SaaS companies in March 2026. The median revenue per employee was $141,125. Bootstrapped companies between $5M and $10M in ARR did best, at $177,240. SaaS Capital, 2026 Revenue Per Employee Benchmarks
So the number I just wrote down is roughly twenty-eight times the median, and I know exactly how that reads.
I have been the guy with the confident number before and I’ve paid for it out of my own pocket. Somewhere around month twelve at my first early-stage startup, I traded six months of salary for equity.
Not as a gesture. As math. The burn rate had a date attached to it, and the date was inside of a year.
We had built the thing. Building was never the problem. We had engineers, a working platform, two large logos, and a founder people believed in. What we did not have was anyone who needed what we made.
That is the part worth sitting with. The build consumed the capital. The market consumed the company. Those are two different failures, and for twenty years the first one was expensive enough to hide the second.
Two things used to be true at the same time. Building was expensive, and adding capacity meant adding people. Both of those stopped being true inside about eighteen months, and almost nobody has gone back and re-run the arithmetic.
The old math was raise, hire, build, and then go find out if anyone wanted it
Think about how this used to work. A founder started with an idea, not a go-to-market strategy. The idea hit, evolved into a vision, and grew from there. Every path looked a little different in terms of hiring, product design, and development sequence, but the shape was the same. You raised capital, you staffed a team, you built.
You did not hire a seller that early. You did not build a pricing model that early. If a sale happened early on, it was founder-led, and it closed on trust in the founder and the vision rather than on the quality of the solution. Those deals came through tight founder networks, not through a funnel.
Then you shipped, showed it to buyers, took the feedback, adjusted, and started grinding it out until you felt the first hint of traction. Faster sales velocity first. Then the best feeling in early-stage company building, which is the day inbound leads start arriving on their own.
The capital raised to build was also the runway you used to figure out the selling. Nobody framed it that way. That is what it was.
And every increment of capacity after that came from a person. More customers meant more account managers. More deals meant more sellers. We called it a hiring plan. It was a subscription to linear growth.
We built a beautiful product for a market that did not exist
I lived this closely at The Big Know in the Twin Cities. I joined as employee seven and the first VP of Sales the company hired. I believed in the solution, the team, and the founder’s vision, and we had early traction with a couple of large companies, which made my move from enterprise media into an early-stage startup feel validated.
Then the early traction stopped being traction and started being two customers.
We had built a complex LMS aimed at high-ticket content marketing. High cost, high complexity, sitting about as far into the upper right as a product can sit on both axis at once. We were in the SaaS graveyard, and we had built ourselves a very nice headstone.
Twelve months in, the choice was pivot or be gone within another twelve. So we tightened the belt, which is where the salary-for-equity trade came in. Anything to slow the burn.
Then came a full year of the actual work. Primary and secondary research. Focus groups. Product design changes. An entirely new approach to marketing to an ideal customer profile we had not yet defined. We did not rebuild the technology. We repointed it at a market that would pay repeatably.
It worked. Traction became something you could feel. Conversations got easier. What we did started resonating, and the market responded. A feeling that is hard to describe. The company was eventually acquired in 2020.
The work, the stress, and the sleepless nights were ultimately worth it. They also burned one thing into my brain permanently: bringing a new solution to market is harder than building it, and it always was.
The expensive half went free and nobody re-priced anything
Fast forward to 2026. Building is no longer the constraint. You already know this, because you have watched someone with no engineering background ship a working product in a weekend. Possibly you were that person.
What has not caught up is everything downstream of the build.
In the past year I have watched founders ship genuinely impressive solutions and then give up. The pattern was consistent enough to be a category. The product got built without needing investment, so they never raised. Never raising meant never buying the runway to learn how to sell. And a founder who is not a seller or a go-to-market strategist defaults to the most expensive belief in software: the product is so good it should sell itself.
It does not. It never did. The difference is that the build used to cost enough to disguise the problem for two or three years.
I am watching the same movie from a different seat
We have been building a business operating system at PerformanceLabs for the past six months, and I am starting to feel the way I felt at my first startup after the pivot landed.
The message is resonating. Revenue and pipeline are building. Time to close is compressing.
“Every company I have worked inside answered growth the same way, by adding a person. That answer was correct for a very long time.”
There are two operators running the company, and we are feeling the pressure. Not the lack of demand pressure. The pressure of spending our hours on work that is not the best use of our gray matter.
So we are contemplating raising a small round, which I have been averse to for two years. The argument that moved me is simple. Why would we cap our revenue at the ceiling of our own capacity?
The unit of the org chart is no longer a person
Here is the part that makes this a genuinely strange moment to be designing an org.
The team structure we’re building is humans plus orchestrated agents, and we have been deliberate about which work goes where. The agents take the repeatable surface: the follow-ups, the status, the drafting, the reconciliation, the thousand small motions that used to justify a headcount. The humans hold the two things agents cannot hold, which are the exception knowledge and the relationship.
That changes the unit of an org chart. The unit is no longer a person. It is a person plus the agents they run. We cap the human count deliberately and let capacity scale underneath it. In our model, AM stands for Agent Manager, not Account Manager.
The internal system we run the company on is the prototype. The codified method is the product. That is the whole bet.
The embedded operator is becoming a go-to-market motion, not a services line item. That is the same conclusion, arrived at from the other direction.
What twenty-eight times the median actually requires
I am not going to pretend the number is modest, so here is what has to be true for it to hold.
Revenue cannot be hours. Five people selling their time caps out long before twenty million, no matter how well they sell it. The mix has to be mostly licensed and platform revenue, with high-touch human accountability as the wedge that earns the right to it.
The method has to be written down, not carried around. Anything that lives only in my head or Eric’s is a bottleneck. If a process cannot be codified well enough for an agent to run it and a human to supervise the exceptions, it will never scale past the two of us.
Every function that grows with the customer count has to be absorbed by the system. Onboarding, reporting, status, renewals, the follow-up nobody enjoys. If a new logo requires a new hire, the model has to be adjusted.
The five seats have to be the five highest-leverage seats. Not five generalists. Five people who each sit on top of a fleet.
And go to market has to be one of the systems, not an exception to them. This is the one that kills the plan if we get it wrong. Five humans cannot personally sell twenty million dollars through a network. That path tops out somewhere well short, which is exactly the ceiling I described earlier.
We are not there. We are two people, a pipeline, and a bet with receipts still being collected. Publishing the number is partly the point. It is harder to quietly revise a target you put in writing.
The Takeaway
If you are a founder or a builder right now, you are probably better at shipping than any founder in history was at your stage. That is not flattery, it is just true. The tooling made it true.
It also made it worthless as a differentiator.
The question is no longer whether you can build the thing. It is whether you can build the machine that puts the thing in front of the right buyer, repeatedly, without you in the room. One-off sales through your network is not a business. It’s just another consulting firm.
So design it. Name the buyer. Name the trigger that makes them care this quarter. Write the pitch down so it survives being said by someone other than you. Price it before the call, not after it. Build the follow-through into a system instead of into your calendar.
We spent a year at my first startup learning that the market decides what you built. You do not have a year anymore. But you also do not need one, because the twelve months we spent repointing a platform is now mostly a Tuesday.
Every company I have worked inside answered growth the same way, by adding a person. That answer was correct for a very long time. It just is not the only one anymore, and the companies still reaching for it first are going to be very confused about why they cannot catch the ones that are not.