At Some Point, Growth Requires Systems, Not Heroes
When I started my business in 2014 with my mother, there were only two of us.
She had her expertise in the kitchen. I had my own skills in management, problem-solving and engineering. At that stage, the company was almost entirely built around what the two of us were capable of doing.
And for a while, it worked extremely well.
The company grew from two people to six, then continued to accelerate. After Covid, our business model evolved and we experienced several years of double-digit growth. At some points, we were hiring almost one new person every month. Eventually, at the peak of the activity, the organisation had grown to around 70 people.
Looking back, one of the biggest mistakes I made during that growth was believing that the management model that had worked for a family business could simply be stretched to fit a much larger company.
It could not.
In the beginning, we operated like a family because, in many ways, we were one. My mother was there. My brother was involved. My wife worked with us. Friends, relatives and people from our close circle joined the company. Decisions were fast because we understood one another. Many things never needed to be written down because they seemed obvious to us.
If there was a problem, I solved it.
If someone needed a decision, I made it.
If an employee was struggling, I stepped in.
When there are five people in a company, this can be an enormous strength. Even with ten people, the founder can still compensate for many weaknesses in the organisation.
But when you reach 30, 50 or 70 people, something changes.
The founder's ability to solve everything stops being a strength and starts becoming a bottleneck.
I remember very clearly the moment when I realised that I had created exactly that situation.
It was a weekday evening. I was at home at around 7 p.m., exhausted, trying to rest. The business was operating seven days a week and, at that time, almost 24 hours a day because some of our production teams worked at night.
I tried to sleep for half an hour.
My phone rang five times.
Every call required a decision.
A production team starting at 10 p.m. had a problem for the following morning. A manager needed approval. Someone else had another operational question. Despite having dozens of employees and several managers around me, important decisions were still finding their way back to me.
I realised that I could not continue like this.
The problem was not only that my employees relied too much on me. I had also trained the organisation to rely on me.
I wanted to know everything. I wanted to understand every issue. I wanted to have the final word because I knew the company inside out and could usually find a solution faster than anyone else.
That works brilliantly until the business becomes larger than one person's mental capacity.
We reached a point where people would call me to ask whether they could replace two broken soup warmers. If we were missing containers, someone would ask me whether they were allowed to buy more. If the electricity failed in one of our three buildings, the issue could still reach me even though each site had its own manager.
Once, a customer was missing an item worth only a few euros and several senior people became involved in deciding what to do.
Today, I would not even want those questions to reach my desk.
But at the time, I had helped create the system that sent them there.
The other side of the problem was equally important. When people did make decisions without me, they did not always make the decisions I expected.
For a long time, I interpreted that as a people problem.
Later, I understood that it was often a management problem.
I knew where the company was going because I had been thinking about it every day for years. I understood what improving margins meant. I knew what we wanted from our environmental strategy. I understood the financial consequences of certain operational decisions.
But many of the people around me did not have that context.
I had made the mistake of assuming that what was obvious to me was obvious to everybody else.
It wasn't.
This became particularly visible when we expanded from one building to three.
In our original location, the culture seemed incredibly strong. We arrived early, left late and worked alongside the teams. If something needed cleaning, we cleaned. If someone had finished their own tasks while another department was struggling, we helped. During quieter moments, we automatically prepared for the following week instead of simply deciding there was nothing left to do.
We believed that this was simply our company culture.
Then we opened another location where no member of the founding family was permanently present.
And suddenly, many of those behaviours disappeared.
Someone who had finished early might simply leave. A driver could finish his job while the kitchen team was still struggling and see no reason to help. If Wednesday was quiet, people did not necessarily think about preparing for the following week or tackling maintenance tasks.
The problem was not that these people were inherently worse.
The problem was that we had mistaken behaviour created by our presence for a system that could reproduce itself without us.
We had never written down that everybody finished together. We had never clearly defined what should happen during quiet periods. We had not sufficiently formalised our expectations around collaboration, quality or decision-making.
Even our product quality exposed the same weakness. In the original kitchen, my mother was constantly present. She tasted things, corrected them and transmitted her standards through daily interaction.
In another kitchen, the chef naturally cooked according to his own interpretation because there was no sufficiently detailed recipe framework telling him exactly what our standard meant.
We were telling people, “We want this to be good.”
But what does “good” mean when the founder is no longer standing next to you?
Growth forces you to answer that question.
A family culture can survive on shared understanding. A scalable organisation needs to translate that understanding into something other people can reproduce.
That transition was painful.
Some employees had joined a small, close family company and genuinely liked having direct access to me. As management layers appeared, they sometimes refused to accept them. They would bypass their manager, ignore HR and insist on speaking directly to me because that was how things had always worked.
At first, I made the situation worse.
I allowed it.
Every time I answered them directly, I weakened the manager I had placed between us. I was asking managers to take responsibility while continuing to demonstrate that the real authority remained with me.
Some employees also experienced the change as a loss of status. Having direct access to the owner had made them feel important. When the organisation became more structured, they no longer had the same privileges or influence.
Some adapted. Others eventually left.
I learned that professionalising an organisation does not mean that every person who succeeded in the previous version of the company will succeed in the next one.
The most difficult lesson, however, was learning to stop being the hero myself.
The first thing I changed was my availability.
When managers came to me with a problem, instead of immediately giving them the answer, I began asking:
“What do you think we should do?”
At first, it was frustrating.
I often knew the answer. It would have taken me thirty seconds to solve the problem.
But every time I did that, I was saving thirty seconds today while guaranteeing another question tomorrow.
So I started listening to their solution first. Whenever possible, I tried to let them use it, even when I might personally have chosen something slightly different.
The second change was clarity.
People needed to know exactly what they owned. Not vague job titles, but real limits.
You are responsible for this area. You can spend up to this amount without asking. These are the decisions you can make. These are the decisions that need to be escalated. These are the results we expect from you.
A manager who has responsibility without authority is not really a manager.
The third change was probably the hardest for me: I had to allow people to make mistakes.
As a founder, you often want everything to be perfect because the company feels personal. When there were only a few of us, I could intervene in almost every detail and push for 100% execution.
But you cannot develop autonomous managers while correcting them every time they deviate from the exact decision you would have made.
This is where KPIs became important.
Instead of fighting over every individual decision, I could define the result.
If a manager achieves a 95% success rate, I have to learn to accept the 5% that did not go perfectly, provided those mistakes are understood and corrected.
Autonomy requires an acceptable error rate.
Otherwise, you do not really delegate responsibility. You simply delegate execution while keeping every decision for yourself.
Over time, we introduced clearer job descriptions, decision limits, annual budgets, weekly management meetings and performance indicators. Operational information became increasingly centralised in our ERP instead of living in people's heads or separate files.
We also became more structured in HR.
Earlier in my career, I often carried too much of my employees' personal situations on my own shoulders. If I knew someone was going through difficulties at home, I might schedule an extra person simply because I was worried they would not cope. If somebody seemed frustrated before a busy weekend, I might change the schedule in advance because I was afraid they would not show up.
I thought I was protecting the company and helping people.
In reality, I was sometimes building an unhealthy comfort zone around problems that had not even happened yet.
There is still something from the family model that I never want to lose: listening.
People should be able to speak. Managers should understand when someone is tired, struggling or going through a difficult period. Good work should be recognised. People should feel respected.
But listening does not mean taking every problem onto the manager's shoulders.
Empathy and accountability can coexist.
The financial consequences of learning these lessons too late were significant.
One expansion project alone ultimately cost us more than €300,000.
What made the growth particularly dangerous was that many weaknesses remained almost invisible while the company was still small. Purchasing worked. Labour costs looked acceptable. Equipment rental seemed manageable. Margins were good.
Then the volume increased.
Because there were no sufficiently strong control systems behind those areas, the deviations grew at the same speed as the business.
Our purchasing margin, for example, deteriorated from around 3.5 to approximately 1.8 at one stage. Tens of thousands of euros disappeared into food purchasing before we reacted quickly enough.
Equipment was being rented because operations needed it, but there was no strong enough financial control saying, “Stop. This is no longer economically rational.”
And when teams were struggling, we kept hiring because it looked like a capacity problem.
We added people on top of processes that were already broken.
Then we added more people to compensate for the problems created by those people.
Operationally, it could still look like growth.
Financially, the model was quietly moving away from the original plan.
That is what makes fast growth dangerous. A small inefficiency multiplied by ten is no longer small.
If I could go back to the moment when we had five, ten or twenty employees, I would put systems in place much earlier.
Not bureaucracy for the sake of bureaucracy.
I would define responsibilities. I would create production and staffing plans. I would formalise operating rules. I would define what quality actually means instead of assuming everybody understands it. I would document recipes and processes. I would establish financial limits and KPIs.
And I would let managers make their first mistakes when the organisation had ten people rather than waiting until it had seventy.
I would also surround myself much earlier with people who had already managed organisations of 50 or 100 employees.
There is no prize for learning every lesson alone.
If an entrepreneur with ten employees told me today, “Why should I create all these systems? Everything works perfectly,” I would probably agree with them.
Everything may indeed work perfectly.
The question is whether it works because the organisation is strong, or because the founder is still close enough to compensate for everything that is weak.
That difference only becomes visible when the company grows.
And by the time it becomes obvious, fixing it can be extremely expensive.
The lesson I took from my own experience is not that companies need more procedures.
It is that a business must learn to operate without depending on exceptional individual effort.
Talent is essential. Great people matter enormously. Founders who care deeply about their company are powerful.
But talent does not scale infinitely.
At some point, growth requires clear responsibilities, measurable expectations, decision rights, management systems and the humility of the founder to stop being the answer to every question.
Because what works because of you at ten people may eventually stop working because of you at seventy.