Every founder who has ever raised money knows the question. How scalable is this, really? How much does it lean on headcount? How many people do you have to add to grow?
It is a fair question. It has been the right question for fifteen years. I just think it is on its way out.
Two kinds of business
For most of my career in tech, there were two kinds of business you could build. SaaS, or services.
SaaS scaled beautifully. Once you have built it, serving a thousand customers costs about the same as serving one. The marginal cost of the next customer rounds to zero, and the margins that follow are the reason a generation of investors fell in love with software.
Services scaled painfully. The output was tied to people, and what a person can do has a ceiling. Every bit of growth meant another hire, and every hire meant more cost, more management, more risk. A services business could be excellent and profitable and still never bend its growth curve the way software did.
So the money went where the scaling was easy. SaaS won the mental model. Services became the thing investors tolerated rather than backed. If your business looked like a consultancy, you learned to describe it as a platform.
The assumption underneath
All of that rested on one thing being true. That a person's output has a ceiling.
It was such a reliable assumption that nobody bothered to state it. It was just how the world worked. One person could handle so many clients, write so many reports, close so many deals, and no more. The ceiling was real, and the whole distinction between scalable software and unscalable services sat on top of it.
That assumption is the thing that is changing.
Then the ceiling moved
When a large part of the work can be directed instead of done by hand, the ceiling moves. Not a little. A lot.
I do not write proposals anymore. I talk to my agents, and together we produce in an afternoon what used to take a week. I am not the exception. Across a lot of knowledge work, one person can now do what recently took a team, because the parts that were slow and manual are handled, and what is left is the judgment that only a person brings.
That does something specific to a services business. The output stops being tied to headcount in the way it always was. You are no longer selling a person's time by the hour. You are selling an outcome, and the software is just the vehicle that delivers it. Some people have started calling this service as software, rather than software as a service, and the flip in wording captures the flip in the model. The deliverable is the service. The software is how it arrives.
A services business built on that does not turn into SaaS. But it stops looking like the old services curve too. It sits somewhere new, with a lot of the leverage of software and a lot of the depth of a real relationship, and I do not think our categories have caught up to it yet.
Why the question is going out of date
The scalability question was a proxy. When an investor asked how much you leaned on headcount, what they were really asking was whether your growth curve would bend the right way, or whether every new customer would drag a new cost behind it.
For fifteen years, headcount was a good proxy for that. If you needed people to grow, you did not scale. If you did not, you did.
The proxy is breaking. A business can now lean on people for the part that matters, the understanding, the judgment, the relationship, and lean on agents for the part that used to require an army. It can grow revenue without growing headcount in lockstep. The old signal, how many people do you add to grow, no longer tells you what it used to.
So the question is not wrong. It is about to have a different answer. Some of the most interesting companies of the next few years might be the ones investors have spent a decade writing off as "doesn't scale." Productize a service far enough, and it starts to behave like software.
Where I could be wrong
I want to be careful here, because it would be easy to overstate this.
I am not saying services now scale like SaaS. They do not. Serving the thousandth customer of a real service still costs more than serving the first, and it always will, because somewhere in it there is still a human who understands the problem and stands behind the result. The gap narrowed. It did not close.
I am not saying headcount stopped mattering. It matters differently. The question is no longer how many people you need to do the work. It is how few people you need to own the outcomes, and that is a different and better question.
And I am not certain. This is a shift I am watching in my own companies, not a law I can prove. The categories are moving, and anyone who tells you exactly where they land is guessing.
But if I had to bet, I would bet that the line between scalable software and unscalable services, the line a whole industry was built on, is getting blurry, and that the people who see it first will build things the rest of the market has been trained to ignore.
That is the shift I am watching.