The Shepherd, The Dog, And The Sheep
A scene from the field that should worry you more than your burn rate I am noticing something weird
A scene from the field that should worry you more than your burn rate
I am noticing something weird when I visit my startup clients. Something right on the verge of exploding, but even my fellow advisors are not seeing it clearly, or it is not discussed enough in public. I do not know. I understand the classic concept of two friends setting up a startup: one tech, one business. Bootstrapping. Working at McDonald’s in the morning and hiding in their parents’ house all night, sculpting their product, excited to be doing something of their own, not even having ten dollars to order pizza. Then six months later, BOOM! An investor blinks in their direction and life changes. They start hiring. A business development team, a technical team, marketing, etc. The company looks complete, giving the product the respect it deserves by their eyes. Then, at market launch, you step back and squint and what do you see? A Shepherd , a Dog, and sheep. No middle management. No connective tissue. No distributed decision makers. Even when there are only ten or fifteen people on payroll, the shape is a straight line with a whistle at one end and a flock at the other.
This scene reminds me of a bigger pattern we have seen before: the shrinking of the middle class in the 1970s and 1980s. Variations of the same story appeared across parts of the US, Europe, Latin America, and later in emerging markets and it is still getting worse till this day. You had a concentration at the top, a swelling at the bottom, and a quiet collapse in the middle that used to stabilize everything. The social economy in general, supply and demand, power of purchase, inflation, and all the related countries or global problems resemble those of a company, small or big. When the middle thins out, society becomes more fragile. When the managerial middle thins out, companies become equally fragile. You can call it poetry or economics. Either way, the ending is predictable.
Let us talk about why founders drift into this structure, the psychology underneath it, why a two-line hierarchy looks efficient but leaks value, and how to rebuild the middle as a strategic advantage instead of a bureaucratic penalty.
The founder’s mind under pressure: protective, heroic, allergic to delegation
New founders, even the rational ones, carry a brain full of alarms. They operate inside a moving blend of pride, fear, scarcity, and identity. You can read every book in the world, but once someone wires a few million into your account and tells you to grow, your nervous system starts negotiating with reality in less-than-ideal ways.
Here is what I see repeatedly in small companies and early-stage startups across the GCC (Gulf Cooperation Council)
Identity fusion with the product. The company is not just something they built. It is them. Criticism of the roadmap feels like criticism of their character. Delegating becomes psychologically expensive because it feels like breaking a piece of themselves and handing it to a stranger.
Scarcity memory. The bootstrapping months program the founder to count every minute and every cent. Scarcity is useful at the start, but it leaves a residue. Even when there is money and people, the founder acts like there is still famine. They hoard decisions because they think hoarding saves waste.
Hero addiction. When the company starts in chaos, being the hero is rewarding. The team watches the founder fix the crisis at 2 a.m., and dopamine writes a love letter to control. The habit sticks. The founder becomes the bottleneck and the brand, the firefighter and the fire alarm.
Illusion of transparency. Founders assume their team understands the full context because it is obvious in their own head. Spoiler: it is not obvious. Without a managerial middle to translate strategy into decisions, people guess. Guessing is not a management system.
Delegation allergy. Delegation looks like trust on a poster. In a real week, it feels like risk. Founders fear the cost of a wrong call more than they value the speed of distributed judgment. So they delegate tasks but not decisions, responsibility but not authority. The team becomes a help desk for the founder’s brain.
Impostor drag. Some founders secretly struggle with impostor syndrome. They would rather keep decisions close than let a manager expose gaps. The tragedy is that the cure for impostor syndrome is actually shared leadership, not solitary perfection.
Cognitive overload disguised as excellence. The founder is always busy. Always in meetings. Always involved. It looks like commitment. Often it is exhaustion. When everything routes through the founder, the company is not scaling. It is aging.
Put those together and you can see why the Shepherd , the Dog, and the sheep model feels comforting. The Shepherd watches everything. The Dog runs around and keeps the herd in a circle. The sheep wait for the next whistle. It is simple. It is fast. It is also the beginning of fragility.
The danger of a two-line hierarchy: fast and brittle beats slow and strong for only a moment
A two-line hierarchy is not the same as a flat hierarchy. A flat hierarchy disperses power and removes unnecessary layers, but it builds capability at the edges. A two-line hierarchy concentrates power at the top and distributes tasks at the edges. That difference is the difference between agility and dependency.
Why a two-line structure fails as you grow
- Decision latency. Every meaningful decision queues at the founder’s door. The backlog grows, the market moves, and you call it “quality control” while outcomes rot in the fridge.
- Context starvation. Individual contributors do not get enough context to make tradeoffs. They optimize locally and accidentally hurt the system. Your engineer saves compute cost, your marketer saves dollars, your salesperson saves time, and you lose the customer.
- Invisible risk. The founders do not hear weak signals because there is no middle to gather and synthesize them. Problems arrive fully grown and angry.
- Talent ceiling. High-caliber people leave because there is no runway to lead. Low-caliber people stay because the job is to follow. You think you are saving on salaries. You are paying in opportunity cost.
- Founder burnout. You can sprint like this for a season. Not for a marathon. I know founders who can outwork a small nation. Every single one of them pays for it somewhere else.
- Culture of hesitation. People learn that initiative gets reversed. They stop trying. You do not need a resignation letter to lose a person. You can lose them while they are still at their desk.
- Board anxiety. Investors are happy when decisions are fast and right. They are less happy when speed is a personality trait, not a system. They know key-man risk when they see it.
A flat hierarchy at its best is not leaderless. It is leaderful. Decisions happen at the level closest to the information. A two-line hierarchy is not leaderful. It is leader dependent. The difference will decide who survives the first real storm.
The neurology of middle management: why the “boring” layer is actually your brain stem
If you enjoy a little neuroscience with your coffee, the analogy writes itself. Think of your org like a nervous system. The executive team is the prefrontal cortex handling strategy, planning, inhibition, long-term goals. The front line is sensory and motor, seeing the world and moving it. Middle management is the brain stem and thalamus, regulating signals, keeping rhythm, switching states, making sure the body does not black out when it stands up too fast. You can live a few seconds without a prefrontal cortex; you cannot live seconds without a brain stem.
Great middle managers translate cognition into motion so the front line executes without guessing, filter noise and escalate signal to protect attention, set a steady cadence with sprints and feedback loops, build memory by documenting decisions, and coach identity by growing people into roles rather than swapping people for them. Calling this “overhead” is like calling your heartbeat a cost center.
The psychology of “no middle”: why it feels efficient and why it is not
It feels clean: one conversation, one approval, instant dopamine, the illusion of speed. Underneath, three biases bite you: present bias (fast clarity today, slower system tomorrow), control bias (believing your call is best because it is yours), and availability bias (choosing what is easiest to recall, not what is most complete). Middle managers widen the lens, add capacity, and de-bias the decision. Learn to love capacity more than control; middle management is not a barrier, it is a multiplier.
Symptoms your middle is missing even if your org chart looks pretty
Team leads with titles but no decision rights, ICs asking permission for local tradeoffs, founders sitting in more than two weekly cross-functional standups, product changes ricocheting between departments, “we move fast” used to justify not measuring, managers spending the week reporting up instead of coaching down, and customer escalations solved by the founder’s phone rather than the team’s system—if you nodded more than twice, your company is pretending to be flat while quietly running a two-line hierarchy with extra Slack channels.
A practical way to integrate a strong middle without killing startup soul
Here is how I advise founders, from Dubai to Riyadh, who want the benefits of middle management without the bureaucracy they fear.
1) Define decisions before you define titles
Write a one-page “decision map” for the next two quarters. List the recurring decisions that make or break your week. Pricing changes, hiring approvals, release gates, partner selection, discount thresholds, incident response, content approvals, roadmap prioritization. Assign each decision a home. Not a person. A level. For example: “Discounts up to X percent live with Sales Manager. Above that, Director of Sales. Exceptions go to CRO.” You just created authority before hierarchy. That is healthy.
2) Create “manager of managers” experiments
Pick two or three high-slope ICs and give them temporary span of control over two roles. Not forever. Ninety days. Provide a checklist of what “good” looks like: weekly one-on-ones, written goals, feedback given, decisions made without escalation. Run it like a pilot. You will learn who wants to lead and who wants a raise without a team. Both answers are useful.
3) Install a lightweight operating cadence
Keep it simple. Quarterly objectives that are actual choices, not wishlists. Biweekly cross-functional standup where decisions are the output, not updates. Weekly team rituals that include a check on decisions made locally. If your meetings do not move authority closer to the work, cancel them.
4) Build a delegation ladder
Teach managers five levels of delegation, from “I decide and explain” to “You decide and I am informed.” Use it explicitly. Put it in job descriptions. Reward managers who climb the ladder for the right topics. This makes authority a discipline instead of a vibe.
5) Protect context like you protect cash
Middle managers need context to be valuable. Share the why, not just the what. Open your board deck to managers. Let them see constraints. People make adult decisions when you treat them like adults.
6) Compensate for judgment, not attendance
Tie part of a manager’s variable pay to decisions kept local and outcomes achieved without escalation. Make “reducing founder involvement” a metric. If you want autonomy, pay for it.
7) Coach the founder out of the center
Set two founder rules for ninety days. Rule one: decline any meeting where you are not the decider or the tie breaker. Rule two: when someone asks for a decision you do not have to make, ask, “What would you do if I were unreachable?” Then approve their answer unless it breaks a guardrail. You are rewiring the culture in real time.
8) Add connective roles, not just controlling roles
Not every manager should be a gate. Some should be bridges. Add program managers, product operations, and enablement leads who exist to reduce friction between teams. Their job is not status updates; it is problem anticipation.
9) Document once, then iterate in tiny loops
Write short decision records. One page. Problem, options considered, choice, why, guardrails. Share across teams. When a similar decision appears, the middle will move faster. This is how you build organizational memory without a religion of process.
10) Design for regional nuance if you operate across markets
In MENA and the GCC, relationship-driven sales and procurement realities often require local judgment. Centralizing every decision in a head office undermines that. Empower regional managers with clear financial guardrails and let them move. You can control brand while letting the market breathe.
None of this kills speed. It stops you from mistaking adrenaline for speed.
The ROI story boards want: numbers, not poetry
Boards and investors want numbers, not poetry, so speak ROI when you make the case for building the middle: cycle time to decision drops when choices are made locally, track it; founder involvement rate falls as more calls get made without escalation, count them; high-performer retention rises when people can grow into real leadership, measure it; customer satisfaction and NPS improve when issues are resolved at first touch, show it; on-time delivery increases when cross-functional tradeoffs happen inside teams, not in the founder’s calendar, publish it.
Objections and the answers you should give yourself
“Middle managers create politics.”
Bad ones do. Good ones create clarity and speed. Hire and teach for the second.
“We cannot afford more managers.”
You are paying for missing managers with founder time, churned customers, and lost talent. Put a number on that, then re-price your belief.
“We will slow down.”
Right now you are fast when the founder is awake and slow when the founder sleeps. That is not speed. That is caffeine.
“We tried managers before and it did not work.”
You tried titles. Try decisions, authority, context, and coaching. Titles without these are costumes.
Culture: the soil that grows or kills the middle
Structure is not enough; culture decides whether authority survives the week. You need psychological safety so people believe they can decide and live to tell the story, explicit guardrails that define the few hard rules protecting brand and risk so freedom exists inside them, curiosity at the top because if the founder always has the best idea the middle will hide their ideas, zero hero worship by celebrating teams that solved without escalation, and public learning where a bad call yields shared lessons instead of blame. Failure is tuition; pay it once.
Conclusion: the middle class of your company keeps the lights on
The shrinking middle class weakens a society because it removes stability, mobility, and the sense that effort creates progress. The shrinking middle in a company does the same. You get volatility at the top, exhaustion at the bottom, and a hollow center where leadership should live. In society, the fix is policy and opportunity. In companies, the fix is design and discipline. Build the middle on purpose. Give it real authority. Measure it by the decisions it keeps local. Your company will feel less like a Shepherd yelling at a Dog to circle the sheep and more like a healthy ecosystem that can cross a mountain without falling apart when the weather changes.
If you want a business that lasts beyond your personal stamina, grow the people who can make decisions you never see. That is not letting go. That is leadership.
Let’s Recap!
Startups love the clean fantasy of a two-line hierarchy, but it is a short road to fragility. Founders resist the middle for psychological reasons that look like efficiency but act like control. The result is decision bottlenecks, context starvation, talent ceilings, and founder burnout. A strong middle does not slow you down. It moves decisions to where the information lives, stabilizes cadence, builds memory, and grows leaders. The fix is practical and cultural: define decisions before titles, pilot manager-of-managers, install a light operating cadence, protect context, pay for judgment, and coach founders out of the center. Do this across regions with local guardrails, and the company scales with integrity. The social lesson applies here too. When the middle collapses, systems wobble. Rebuild the middle, and the whole organism gets smarter, faster, and harder to break.
