My previous co-founder and I started our digital agency in 2009. We had met a few years earlier at a large international advertising agency and instantly clicked. He was an animator working mostly in Flash at the time. I was a web developer focused on the backend. We were both freelancing on the side and having more fun working together than working alone. Starting something jointly felt obvious.
Neither of us had a plan beyond that. We knew we worked well together, we shared a sense of what good digital work looked like, and we both had a discipline the other one did not. The whole foundation of the agency was: we make good things, and we make them together.
For the first few years, that was enough.
Then, slowly, the work changed shape. We did not.
Four jobs wearing the same title
What nobody told me when we started was that running an agency is not one job. It is a sequence of jobs that look similar from the outside and feel completely different from the inside. The job rotates roughly every three to four years, and each rotation asks a different version of you to show up. The earlier versions of the job do not prepare you for the later ones. In some cases, they actively unfit you for them.
The first job is the craft job. Two people, sometimes three, doing the work directly. You and your partner are the production engine. The agency's reputation is whatever you ship that month. The conversations are about the work itself - what is good, what is not, how to make this thing better. Both partners are in the same room, on the same project, solving the same problems. The partnership is strongest here because there is almost no distance between the work and the relationship.
The second job is the team job. You hire people. The agency becomes ten, then fifteen, then twenty. The partners stop building most things themselves. The work is still close - you can still see every project, sit in every important meeting, review every output - but you are no longer the producer. You are the standard. Your job is to define what good looks like and make sure the team can produce it without you. Some partners love this job. Some find it strange and a little lonely. It is the first version where the partners can begin to want different things without noticing.
The third job is the business job. Twenty people becomes thirty, sometimes more. The agency now has its own gravity. Cash flow, hiring, leadership, positioning, partnerships, contracts, operations, infrastructure. The work is happening, but you do not touch it anymore. You touch the system that makes the work happen. Some days you do not think about a project at all. You think about retention, about margin, about the next two hires, about whether the team you have can carry the next phase. The partnership lives here, too, but the conversation has completely changed. You are no longer two craftspeople comparing notes on a build. You are two operators comparing readings on instruments - and the readings often disagree, because each of you is interpreting them through a slightly different sense of what the agency is now for.
The fourth job, if you get there, is the leadership job. You are running a business that is mostly run by other people. Your time is spent on direction, on culture, on the things only you can do, on the conversations only you are in. The team has its own rhythm. Clients are managed by leaders you trust. The thing you built has stopped being yours in the way it used to be. It belongs to the people running it day to day. Your job is to make sure it keeps belonging to them, and to point it somewhere worth pointing.
These four jobs are not separate companies. They are the same company, expressed differently at different sizes, demanding different things from the people at the top.
And almost no founder partnership renegotiates itself at each rotation.
Where ours came apart
What I now think happened with my co-founder and me is that we were perfectly fit for the first job, and quite well fit for the second. By the third, we were diverging. By the time it was clear we were diverging, we had already been doing the third job together for years, and the agency had grown around the diverging shape of our partnership without anyone naming it.
The drift looked like a lot of small things. We made different calls. We prioritized differently. We had different instincts about which clients to chase, which hires to make, which projects to refuse. We had different relationships with the business itself. He said often that he enjoyed building the company more and more. I think he truly did. In hindsight, I do not think I ever really did. I had convinced myself I did because the role required it.
Do not get me wrong, I loved the people we hired, enjoyed working with most of our clients, admired the work we did, and I am still very proud of what we accomplished as a team. I think what I struggled with was no longer being able to create things myself. Doing administrative work, creating proposals, visiting clients, chasing invoices, worrying about cash flow - all of that is just different. And it was never really my thing.
The agency did not need both of us to enjoy the same things. It needed both of us to be honest about which things we enjoyed, and to organize the partnership around that honesty.
We were not honest with each other in that way. Not because either of us was lying. Because the conversation was not happening at all. The work was busy. The team was growing. Clients were calling. The kind of conversation that would have surfaced the drift - a slow, quiet, slightly uncomfortable one - kept getting deferred to a calmer week that never quite arrived.
By the time it did surface, the drift was not a difference of preference. It was a difference of direction. And differences of direction inside an agency tend to be irreversible.
The pattern is not particular to us
I am not the first founder to live through this. I have watched it happen in other agencies, often in slow motion, often with founders who respect each other. The pattern is usually the same.
The partnership was built on the shape of an earlier job. The job changed. The partnership did not. By the time anyone names the gap, the gap has been doing its quiet work for years.
None of this happens because founders are bad at partnerships. The agency is constantly remaking the founders, in different directions, faster than the partnership renegotiates itself. One founder gravitates toward the business mechanics. The other gravitates toward the work. One enjoys leading the team. The other enjoys building the thing. One wants to grow. The other wants to deepen. None of those are wrong. They are just incompatible expressions of two people who used to want the same thing because they were doing the same thing.
The longer the partnership runs without an explicit renegotiation, the more each person calcifies into a version of themselves that the partnership cannot hold. Eventually, somebody leaves. Or worse, nobody leaves and the agency keeps running with the cost of the mismatch baked into every decision.
What would have had to be true
What I think about most is not the breakup itself. It is what would have had to be true for the partnership to keep working.
It would have meant naming the rotation as it happened. "The job is changing. Let's talk about what it is now and which parts of it each of us wants to do". It would have meant being honest about what each of us actually liked, not what we thought we were supposed to like. It would have meant restructuring the roles before the dissatisfaction became visible, rather than after. It would have meant treating the partnership itself as a thing that needed maintenance, like any other system inside the agency.
We did none of that. Neither of us thought to. We were busy.
The thing that kills founder partnerships is rarely conflict. It is the slow accumulation of unspoken differences that never got the conversation they needed, because the conversation always felt like it could wait one more quarter.
The system underneath the system
This is part of what Peter, my business partner, and I have been thinking about as we build ClockWork League. Most of the work agencies need is structural - operating systems, clarity on positioning, real cash discipline, a way of leading the team that does not depend on the founder being in every room. But underneath all of that, there is a quieter layer that nobody puts in an agency-growth program. The founders themselves are being remade by the business every few years, in different directions, often without noticing.
Agencies that scale well are agencies that catch the rotation as it happens. The founders renegotiate their roles deliberately. The partnership is treated as a relationship that needs work, not a foundation that automatically holds. The conversation about what each person actually wants from the next version of the agency happens before the drift becomes a direction.
That is harder than building any system inside the business. It is also the part most agencies skip.
I am still proud of what we built. I do not regret starting it. I learned more in those twelve years than in anything else I have done, and most of what I now think about agency growth was shaped by that experience, including the parts that ended badly.
But I understand now that the partnership was the system underneath the system. We kept tuning the agency and we never tuned the partnership. The agency outgrew the shape of the people running it, and the people running it did not know how to grow their relationship at the same speed as the company.
If I were starting an agency with a partner today, the first system I would build is not a positioning framework or a delivery process or a sales pipeline. It is a recurring, scheduled, deliberate conversation between the partners about what the job has become, what each of us wants from it, and what we need to change before the next rotation arrives.
Nearly everything that broke our partnership would have surfaced in that conversation a year earlier than it did.
The agency would still have changed. The job would still have rotated. But the partnership would have had a chance to rotate with it.