I have learnt that a founder can remain the bottleneck after signing every delegation document the company needs. The problem is not always an unwillingness to let other people decide. Sometimes the founder enjoys being helpful, answers a question in passing and fails to notice that the conversation has displaced someone else's authority. Colleagues learn that the formal route is provisional and that a private appeal may produce a quicker or more favourable answer. The organisation then carries two systems: the one it publishes and the one it trusts under pressure.
In Wensharo's early years, direct access to me was useful. I knew many customers, understood the equipment and could bring several considerations together quickly. The team was small enough for a conversation to reach most people affected. Growth changed those conditions. My answer might no longer include the latest service evidence or the commitments another manager had already made. Yet the history of the company made my opinion carry more weight than the quality of the evidence available to me at that moment. Experience and ownership were being mistaken for current context.
Notice the route people actually use
When we reviewed engineering and spare-parts approvals, the median elapsed time in a defined cohort was nineteen working days. We could have responded by asking managers to submit clearer papers or by allocating more time in my diary. Both might have helped at the margin. Neither addressed why those decisions were reaching me. Twelve managers had enough competence to own much of the work, but the organisation had not made their authority dependable. A disagreement could still travel around them and return as an instruction whose origin nobody wanted to challenge.
We began by specifying the decisions each manager could make, the relevant financial and technical boundaries and the circumstances requiring escalation. This was not an attempt to remove every ambiguity from industrial work. It was a way to make ordinary responsibility clear enough that an unusual case could be recognised. Urgent service matters continued through their established response process. The change concerned the routine queue in which people were waiting for founder reassurance even when the issue belonged within an existing managerial remit and the necessary evidence was available to that manager.
The next step was to examine bypasses, including my own. If someone asked me to revisit a decision, I needed to know whether the responsible manager was involved and whether new information justified reopening it. Referring the question back could feel less helpful than giving an immediate answer. Long-serving colleagues sometimes heard it as a reduction in their relationship with me. I had to explain that personal access remained possible while decision authority had become more explicit. The company could not scale a process in which trust was demonstrated by obtaining an exception from the founder.
Let a manager make a different defensible choice
Delegation becomes difficult when another capable person reaches an answer I would not have chosen. It is easy to support managerial authority when it reproduces my preference. The real test is whether the reasoning fits the agreed remit, uses the available evidence and accepts the consequences. I still have a responsibility to challenge a weak decision. I do not have a right to label every difference as weakness merely because my experience leads me towards another option. A team that must guess the founder's answer before acting has received responsibility without meaningful discretion.
Two decisions in the reviewed programme were later reversed after additional information emerged. We examined those reversals openly. They did not automatically prove that delegation had failed, nor did the new structure excuse poor work. We asked what had been known, what should reasonably have been checked and what changed later. This distinction matters because industrial decisions are rarely made with complete information. If every adverse development leads to authority being withdrawn, managers will protect themselves by escalating more, and the founder will recreate the queue while blaming the team for its lack of confidence.
Over nine months, median approval time in the defined group fell from nineteen to six working days. I regard that as useful evidence of a changed process, not as a complete measure of organisational maturity. Faster decisions can still be poor decisions. We therefore retained review of consequences and of cases that crossed the agreed boundaries. The reduction mattered because it occurred alongside a clearer account of who owned the work. It would have been less valuable if we had simply encouraged managers to move quickly while leaving an informal founder veto available whenever someone disliked the outcome.
There is a reasonable objection that founders sometimes see a connection others miss. Their experience, customer knowledge and willingness to carry risk may be central to the business. Delegation should not silence that contribution. I can offer a perspective, ask a question or require escalation when the matter genuinely belongs with the chief executive or board. The important point is to make the capacity in which I am acting clear. A suggestion should not become an invisible instruction, and an instruction should not arrive disguised as a casual comment that nobody feels able to question.
Separate ownership from the next operating answer
The minority investment and separate chair helped me see how easily my roles could blur. As majority shareholder, I have an interest in the long-term value and direction of the business. As a director, I participate in oversight and major decisions. As CEO, I lead the organisation within its agreed framework. These perspectives are connected, but they do not give every personal preference the same standing. The chair has sometimes asked whether I am raising an enterprise issue or reopening an ordinary operating choice because I remain accustomed to being the person who decides it.
I would not prescribe the same delegation pattern for every founder-led manufacturer. A company facing an exceptional event may need a more concentrated response for a defined period. A manager may need development before a broader remit is appropriate. The boundaries should reflect competence and circumstances. What should not happen is a temporary intervention quietly becoming the permanent way business is done. I want any change in authority to be understood, communicated and revisited, so the company can distinguish a necessary response from a return to familiar habits that no longer fit its size.
The work remains personal. I have to tolerate a period in which answering fewer questions can feel like contributing less, even though the organisation is becoming more capable. I also have to remain accountable for the people, information and review process that make delegation responsible. The aim is not a founder who disappears. It is a company that can use the founder's judgement without making every judgement depend on him. That requires closing the private appeal court as carefully as opening the formal decision route, and accepting that the change begins with the person who once made direct access useful.