The Three Decisions Every COO Must Make Before the Company Reaches 80 People
Your CEO has stopped “letting decisions out.” Not because she is lazy, but because she has become a bottleneck.
At 50 people she still thought she could know everything and approve everything. At 70 she is overwhelmed. At 80 the company enters what researchers call execution paralysis: decisions queue up, hiring slows down and the best people leave because there is no more room to move.
It happens to most founders between 50 and 100 employees. The processes that worked at 30 no longer work at 80. The organizational structure collapses quietly. And by the time the CEO notices, key decisions remain pending for weeks.
Research shows that decision-making bottlenecks can drain up to 30% of revenue and 26% of productivity. It is not a “small” problem. It is an existential problem.
And here comes the hard part: you can’t simply work harder or hire more people to fix it. The point is not effort. The point is organizational structure. You need three critical structural decisions, and you must make them before you reach 80 people, not after.
Why the 50–80 Person Threshold Is the Critical Turning Point
Most growing companies go through predictable organizational breaking points: at 10 people, 25 people, 50 people and 80 people.
At 10 people, it’s no longer possible to operate only with informal communication. Someone must start taking notes. Early processes emerge because chaos becomes unsustainable.
At 25 people, the first level of management is required. The founder can no longer directly manage 25 people. Delegation becomes mandatory.
At 50 people, however, something different happens: the founder’s decision architecture collapses. This is the true critical turning point that almost no one sees coming.
Why? Because at 50 people the company already has such a level of complexity that decisions require input from multiple functions (Product, Engineering, Sales, Operations). But there is not yet a sufficiently solid organizational structure to route these decisions efficiently. So everything rises upward, back to the CEO, who becomes a human bottleneck.
At 80 people, this bottleneck becomes fatal. Execution speed drops 30–40% compared to the 30–50 phase. Good people leave because they are frustrated by the constant waiting for decisions. New hires enter a culture of waiting, not building.
The window to intervene is narrow: between 50 and 80 people.
If you make the three right structural decisions in this interval, you can scale smoothly beyond 150 people.
If you miss it, you risk spending 6–12 months in organizational limbo, often forced into a deep restructuring that costs on average 15–20% of annual productivity.
This is not an issue of growth, but an issue of decision architecture.
Decision 1: Clarify Decision Rights
The first structural decision is almost always misunderstood. It is not about who has authority, but about the speed of decisions and their reversibility.
Type 1 decisions (irreversible, high impact)
These decisions must go through the CEO. They are difficult to reverse and have significant strategic impact: key hiring or firing, meaningful changes in product direction, strategic partnerships, major budget allocations.
Type 1 decisions should be made methodically: involving multiple stakeholders and only after thorough analysis. In a healthy company they should represent about 10–15% of all decisions.
Type 2 decisions (reversible, local impact)
These decisions should be made by whoever is closest to the problem and possesses the specific expertise. They are easily corrected if wrong. In most well-functioning organizations, the vast majority of decisions are Type 2.
Type 2 decisions must be made quickly. Managers must have the authority to make them without CEO approval. This is what creates decision velocity: the organization’s ability to move, adapt and execute without unnecessary blocks.
Explicitly clarifying this distinction does not reduce the CEO’s control. On the contrary, it frees up decision time for the few truly critical choices and unlocks execution on everything else.
The Boveda Case
Boveda, a U.S. humidity control systems manufacturer based in Minnesota, encountered exactly this problem when it reached 47 employees. The company was doing $25M in revenue and had a very engaged team growing at an average of 33% per year. Yet growth had begun to slow.
The main cause was clear: the leadership team (6 people) had become a bottleneck.
In particular, the decision-making process had been designed for a small startup. Everything passed through the CEO. With increased business complexity, this approach generated delays, frustration and poorer-quality decisions because the right perspectives were not present when decisions were made.
The solution adopted by Boveda was structural, not tactical:
Clarify decision types: they defined which decisions were Type 1 (requiring CEO approval) and which were Type 2 (teams could make them autonomously)
Expand the leadership team: from 6 to 13 people, including functional leaders who previously had never “had a seat at the table”
Document the framework: making it visible and training managers on its use
The effect was immediate. As they internally reported:
“Previously, decisions already made regularly returned to the leadership team’s weekly agenda because new information emerged that was not present at the time of the first decision. The new process based on Type 1 and Type 2 decisions gave the entire company a shared framework for decision-making.”
Within two years, Boveda went from $25M to $50M in revenue and from 47 to 82 employees, nearly doubling in size without losing execution speed.
This case demonstrates a key point for those leading growing companies: it’s not lack of talent slowing the organization, but a decision architecture that doesn’t scale.
Decision 2: Build Management Levels
Your CEO should have 5–7 direct reports. Not 15. Not 20.
It sounds simple, but most growing companies don’t do it. Why? Because founders resist hierarchy. They think more layers mean more bureaucracy. They want to “stay connected to the team.”
It’s a mistake. And a costly one.
Research on span of control
McKinsey data shows that when a leader manages more than 7–8 direct reports, two things worsen simultaneously:
the quality of decisions
the effectiveness of delegation
The leader doesn’t have enough time to make thoughtful decisions about each person, and managers don’t have sufficient access to the leader to decide well themselves.
The optimal span of control is not about how “flat” the organization is. It is about the managerial depth needed so that decisions are made at the right level without repeatedly rising upward to the CEO and creating bottlenecks.
When to add levels
Between 50 and 80 people, it is typically necessary to introduce the second (and in some cases the third) level of management:
Level 1: CEO + direct reports (5–7 people)
Level 2: functional heads / department heads (15–25 people)
Level 3: team leads / senior individual contributors (40–80 people)
This structure does not add bureaucracy. It adds distribution of decisions.
It enables faster decisions because decisions are made closer to the problem instead of accumulating on the CEO’s desk.
Which roles to add
Most companies between 50 and 80 employees need to introduce:
VP of Operations (if not already present): responsible for processes, systems and scalability
Director of Product: responsible for product roadmap, priorities and cross-functional coordination
Director of Engineering: responsible for technical architecture, team structure and execution
VP of Sales (if revenue-driven): responsible for sales process, team structure and quota management
The exact roles depend on the business model, but the principle is always the same: create functional leaders who own their domain, can make Type 2 decisions autonomously and have direct access to the CEO.
A common mistake: hiring for title, not capacity.
Most companies make this error when adding levels: they hire a VP of Operations who has never scaled an organization, and then wonder why execution does not improve.
You need people who have already crossed the 50–100 employee phase. People who know the dynamics, know what breaks and know how to build scalable processes.
Decision 3: Implement Cross-Functional Processes
At 30 people, everyone knows everyone. Communication happens at lunch. If Engineering has a technical problem, they go to Product and talk about it. If Sales has a customer request, they pull Product aside and discuss it.
At 80 people, however, departments exist. Product doesn’t talk to Sales. Engineering doesn’t talk to Support. No one shares common goals, so everyone optimizes for their department metrics, not for company results.
Silos don’t form because people are bad collaborators; they form because the organizational structure naturally generates them.
Without intentional cross-functional mechanisms, compartmentalization becomes the default state.
What breaks without cross-functional processes
When structured collaboration mechanisms are missing, problems are predictable:
Product launches take up to 3× longer because requirements and priorities are not aligned between Product and Engineering
Support issues drag on forever because Support and Product do not share common metrics
Sales promises what Engineering cannot deliver because Sales and Product don’t talk
Strategic changes are never executed because each function optimizes its own objectives, not shared outcomes
The result is not just inefficiency, but systemic frustration.
How to implement cross-functional processes
The right time to build them is now, not when silos have already solidified.
1. Shared OKRs (not departmental)
Instead of having separate OKRs for Product, Engineering and Sales, create company-level OKRs that require cross-functional execution.
Every function contributes to the same outcome through its own levers.
Example:
Company OKR: “Increase customer retention from 92% to 96%”
Product: reduce onboarding friction (measured by time-to-first-action)
Engineering: improve system reliability (measured by uptime)
Support: reduce ticket response time (measured by MTTR)
All functions work on the same company objective but with different levers. This creates alignment without micromanagement.
2. Weekly cross-functional meetings
Product, Engineering, Sales, Support and Operations meet for 60 minutes weekly to:
discuss major blocks
escalate cross-functional decision items
celebrate results
align on priorities for the next week
This is not a status meeting; it is a coordination meeting. It prevents surprises and intercepts misalignment before it becomes costly.
3. Joint decisions on strategic initiatives
When launching an important feature, modifying a key process or introducing a strategic change, involve all functions from the beginning.
Don’t let Product decide alone and then “throw it over the wall” to Engineering.
Decisions made together travel faster and break less along the way.
Implementing cross-functional processes doesn’t mean slowing the organization. It means making execution possible when complexity grows.
At 80 people, collaboration is no longer spontaneous: it must be designed.
The fintech case
A fintech company introduced a “collaboration by design” initiative, designed specifically to address tool fragmentation and the creation of process silos across departments.The solution was the creation of a unified technical platform, integrating design assets, code repositories, and product documentation.
The results were measurable and immediate:
–50% of issues in the design-to-development handoff Designers and developers no longer worked in isolation. Specs, assets, and feedback flowed continuously and shared.
–45% of clarification meetings on requirements When everyone has access to the same documentation, fewer meetings are needed to understand what is actually required.
+60% documentation completeness Cross-functional processes force better documentation, because more teams depend on the same materials.
+30% faster onboarding speed for new hires New joiners integrate faster because they can immediately see how cross-functional workflows actually work.
McKinsey data confirms the picture: organizations with strong cross-functional collaboration are 1.5 times more likely to achieve above-average growth.
Not slightly above average. Above average.
And it is a difference that, over the long term, radically changes a company’s trajectory.
Common Scaling Mistakes
Mistake 1: Hiring leaders for title, not scaling experience
In this phase you need people who have already been through it. A VP of Operations who has only managed teams of 20 will struggle at 80. They have not seen the typical patterns. They don’t know what breaks when complexity increases.
How to avoid it: When hiring leadership between 50 and 100 people, prioritize those who are proven at scale, not those with fancy titles.
Mistake 2: Considering organizational levels as bureaucracy rather than decision distribution
Many founders resist new levels because they fear bureaucracy. But levels are decision distribution, not bureaucracy. Levels allow decisions to be made faster and better because they are made closer to the problem.
How to avoid it: Frame new levels as “decision distribution,” not “management layers.” Measure success by decision speed, not number of meetings.
Mistake 3: Creating shared objectives without shared accountability
You can create as many cross-functional OKRs as you want. But if there are no mechanisms of shared accountability, functions will continue optimizing for their own objectives.
How to avoid it: Make cross-functional success part of variable compensation or individual bonuses. If Sales’ bonus depends only on closed revenue, the team will ignore retention (which requires collaboration with Support). When compensation aligns to shared objectives, behavior changes.
Mistake 4: Not documenting decision rights
You can have the best decision framework in the world, but if it exists only in your head (or the CEO’s), it doesn’t scale. Managers will make decisions inconsistently. People won’t know who decides what.
How to avoid it: Document the decision framework in writing. Make it accessible. Update it as the organization evolves.
Mistake 5: Scaling processes before scaling structure
Many companies try to introduce better processes (OKRs, cross-functional meetings, alignment rituals) before fixing structure. But processes only work if the structure supports them.
How to avoid it: Structure first, then processes. Clarify decision rights and span of control before implementing processes and frameworks on top.
These mistakes don’t just slow growth. They make the organization fragile at the very moment it should become stronger.
Metrics to Monitor Progress
Once you’ve made these three decisions, how do you know if they are actually working?
Here are the metrics that matter — the ones that tell you whether decision architecture is unlocking execution or not.
Decision Velocity
Average time for Type 2 decisions: how long it takes to make a decision that doesn’t require CEO approval?
Target: 1–3 days.
If it takes one to two weeks, the decision framework is unclear or not respected.Average time for Type 1 decisions: how long to make a decision requiring CEO approval?
Target: 5–10 business days (excluding time for discussion and analysis).
If it takes more than three weeks, the CEO is overloaded or the scope of Type 1 decisions is too broad.
Managerial Autonomy
Percentage of decisions made by managers without CEO approval.
Target: 70–80% of all decisions.
Below 60% means the bottleneck is still there.Manager satisfaction with decision authority.
Ask directly via survey:
“Do you feel genuinely empowered to make decisions?”
If the answer is no, the framework exists on paper but not in practice.
Execution Speed
Feature delivery time: how long from idea to release? Track month over month. After fixing structure, this number should improve.
Time-to-hire: how long from decision to hire to the first day on the job? This indicator often improves significantly when management levels are introduced because hiring decisions are distributed to functional leaders instead of centralized on the CEO.
Organizational Health
Voluntary turnover: do your best people stay? If you lose strong talent 6–12 months after surpassing 50 people, it is often due to frustration linked to decision bottlenecks.
Cross-functional collaboration: measure via periodic surveys. Do people perceive that Product and Engineering work together? Or do they still feel silos?
Engagement: engagement tends to drop in the 50–80 person phase. Your goal is not just to halt the decline, but to stabilize or improve engagement while implementing these changes.
Growth Metrics
Revenue growth rate: often slows precisely in the 50–80 person phase. When the decision bottleneck is removed, growth should regain momentum.
New customer acquisition: follows a similar pattern to revenue. Execution paralysis often manifests here first: CAC increases or acquisition pace slows because the organization cannot move fast enough.
These metrics are not for reporting, but to answer one critical question: is your decision architecture accelerating execution or slowing it? If you measure these dimensions continuously, you will know long before the problem becomes costly.
Implementation Roadmap
You don’t need to implement all three decisions at once. The key is the sequence. Here is a practical and sustainable path.
Month 1–2: decision rights
Start here. It’s the fastest part to implement and has the highest immediate impact.
Key actions:
Clearly document Type 1 and Type 2 decisions
Train the leadership team on the decision framework
Begin actively delegating Type 2 decisions to managers
At this stage you are not changing people or roles; you are changing how decisions are made.
Month 2–3: hire the next functional leader
Now that decision rights are clear, you need a structure that supports them.
It’s time to hire the first new functional leader — often a VP of Operations or a Director of Product. The choice depends on where you feel the bottleneck is strongest today.
Month 3–4: add the second level of management
After the first hire, begin building the second level. Depending on context, you can hire multiple roles in parallel or spread them over 2–3 months. The goal is not to fill the org chart, but to distribute decisions sustainably.
Month 4–6: cross-functional processes
By this point you have a stronger leadership structure. It’s the right time to introduce:
Shared company-level OKRs
Weekly cross-functional meetings
Joint decision rituals for strategic initiatives
These mechanisms work much better when the right people are already at the table.
Continuous: measure and iterate
Start monitoring the metrics described earlier. Each quarter:
Evaluate what is working
Identify what is slowing the organization
Adapt the framework based on the real evolution of the company
Decision architecture is not static: it must evolve with organizational complexity.
Conclusions
The transition from 50 to 80 people is one of the most critical turning points in a company’s life. The structures that got you to 50 stop working. The CEO becomes a bottleneck. Silos form. Execution slows.
But it doesn’t have to be that way. If you make three critical decisions (clarify decision rights, build management levels with appropriate span of control, and implement cross-functional processes) you can navigate this phase smoothly.
In fact, you can accelerate exactly when others get stuck.
The cost of making these decisions: a few weeks of structural clarity and some difficult conversations about decision rights and accountability.
The cost of not making them: months of execution paralysis, organizational frustration and loss of key people.
The math is obvious.
Are you between 50 and 100 people? Is your CEO becoming a bottleneck? Is your organizational structure beginning to buckle under the weight of growth?
This is exactly the moment when these decisions make a difference.
I have helped founders like you navigate this transition. Those who move quickly, make clear decisions and implement them consistently manage to scale past 150 people without friction. Those who delay or apply these decisions only halfway often remain stuck for 12–18 months.
Fill out the form below to start a conversation about how ready your organization’s decision architecture is for the next stage of growth.
No pitch, just an honest assessment of what works, what is breaking, and what needs to be fixed before it becomes costly.
I am Francesco Malmusi, founder and former C-level operator. I help CEOs, founders and COOs manage growth moments like this — when the structure that got you to 50 people no longer works and you need to rethink how decisions are made, how leadership is organized and how the company actually executes.
If you find yourself exactly in this situation, let’s talk.
