Your company cannot scale if every decision needs you

You approve a customer discount between two interviews. Then a designer needs a decision, operations needs an exception, and someone asks whether the new landing page is allowed to say “simple.” Meanwhile, an important partnership discussion slips again. Everyone is working. Everyone is waiting for you.

A founder bottleneck can develop for respectable reasons. You know the customers, carry the financial risk and remember why earlier ideas failed. Your judgement helped build the business. As the company grows, however, routing ordinary decisions through one person adds a limit that recruiting alone cannot remove.

Find the queue before judging the people

For one working week, keep a decision diary. Record what reached you, who brought it, how long it waited, why they could not decide and what happened after your answer. Include interruptions in messages. Your calendar may show three approval meetings while your phone contains the other thirty.

Separate useful involvement from accidental dependency. A major investment may genuinely belong with you. Choosing between two already approved email layouts probably does not. The informative question is why the second decision still arrived. Perhaps the standards are unclear. Perhaps people were criticised last time they made a reasonable call.

Ask team members: “Which decisions would you feel comfortable owning if the boundaries were clearer?” This invites useful detail. “Why are you all so dependent on me?” mostly invites an excellent view of the meeting-room carpet.

Give people a decision, a boundary and the relevant context

Delegating “marketing” is too large to be useful. Delegating the next campaign's choice of audience, within an agreed budget and positioning, is something someone can actually own. Define the result, the budget, the constraints, the information required and the situations that trigger escalation.

Imagine a small subscription business as a hypothetical example. Its founder approves every customer credit. The support lead could own credits up to an agreed amount where a documented service failure exists. Unusual contractual commitments, repeated abuse and larger amounts would still be escalated. The exact thresholds require the company's own financial and operational judgement.

Now support can act and the founder retains a clear view of risk. The policy needs examples of ordinary and exceptional cases. A threshold without examples can simply move the queue to the question “Does this count as ordinary?”

Separate advice from approval

People sometimes ask you for a perspective, then interpret your perspective as a command. Tell them which conversation you are having. “I am giving input here; you own the final choice” is a small sentence with a large practical use. Conversely, when you retain the decision, say so before the team spends a week pretending otherwise.

Atlassian's DACI framework distinguishes the person driving a decision, the approver, the contributors and those who need to be informed. You do not need to import its entire ceremony. The useful principle is to make these different jobs visible rather than treating every interested colleague as an additional approver.

For recurring decisions, write the arrangement once. For expensive or unusual decisions, agree it explicitly each time. Adding an acronym to an unclear process will otherwise create an unclear process with branded stationery.

Calculate the cost of waiting without pretending it is exact

Use a modest illustrative calculation. Assume eight decisions each week leave two colleagues unable to continue for ninety minutes. That represents 24 person-hours of blocked capacity: eight times two times 1.5. This is a scenario, not a measured result or a promise that all those hours become revenue.

Some people will switch to other useful work. Others will incur extra coordination or lose a timely opportunity. Compare the calculation with what actually happened in your diary. The commercial question is which delays affected a customer, a launch, a sale or a decision that mattered.

Also count your own attention. Saving ten minutes on a minor decision may create room for a difficult conversation that only you can lead. Treat that room as capacity to allocate deliberately, rather than as an invitation to accept another meeting about the logo.

Stop taking the decision back halfway through

A manager cannot learn to own a decision if the founder reverses it whenever a different preference appears. Agree review points and acceptable risk beforehand. If the choice falls within those boundaries, allow a different reasonable answer. Ask what the owner considered and what they expect to learn.

This does not require watching a preventable serious mistake unfold. Intervene when a material constraint is missed, new evidence changes the situation or a risk exceeds the agreed boundary. Explain the reason. “I prefer the other option” and “this exposes information we must protect” are different grounds for intervention.

If decisions repeatedly disappoint you, inspect the inputs. Did the person receive customer context, cost information and access to the right specialists? Accountability is much easier to exercise when someone can see the problem they are accountable for.

Choose your first delegation experiment

• Select one frequent, reasonably reversible decision.
• Name one owner and identify the people whose input matters.
• Write three examples of acceptable decisions and two escalation cases.
• Agree a review date and a simple record of decisions made.
• Tell the wider team that approval now sits with that owner.
• Review speed, quality and surprises together before widening the scope.

Keep the experiment small enough to discuss honestly. Start with scheduling, routine proposals or bounded customer exceptions rather than the most consequential decision in the business. A successful first transfer gives both sides a concrete experience to build on.

Make your continued involvement useful

The goal is a founder whose attention reaches the decisions that justify it. You may still lead vision, capital allocation and a handful of important customer relationships. The team should understand where your involvement makes the work better, and where it merely makes the work later.

At the review, try: “Which boundary helped? Where did you need context I had not shared? Which decision should move next?” That conversation develops judgement on both sides. Delegation changes the founder's job too.

If the queue exposes a wider leadership gap, read what to clarify before hiring a Head of Product. If everyone is also juggling unfinished work, the busy-team diagnosis helps separate approval delay from overload. I help leadership teams map decisions, clarify product direction and improve delivery. You can bring me one real decision that keeps getting stuck and explain who is currently waiting.

Questions people ask

How do I know if I am the founder bottleneck?

Track the decisions that wait for your approval for a week. Look for repeated, reasonably reversible choices that another person could own with clearer context and limits. Ask what prevented that person from deciding. A founder bottleneck is visible in avoidable waiting and repeated escalation, rather than simply in a busy calendar or a large number of direct reports.

How can a founder delegate without losing control?

Delegate a specific category of decisions within agreed financial, customer and operational boundaries. Name the owner, define escalation cases and review the actual decisions at a set interval. Retain the decisions whose risk genuinely requires your involvement. Control comes from visible responsibilities and useful feedback, rather than from approving every ordinary choice before anyone can act.

Which decisions should a founder delegate first?

Start with frequent choices that are reasonably reversible and whose consequences the team understands. Examples might include scheduling, routine proposal preparation or bounded customer exceptions. Avoid beginning with an unfamiliar, high-risk commitment. Provide examples and relevant information, then review quality and waiting time together before expanding authority to more complex decisions.