Eat What You Kill
The org chart is the enemy
Business has never been more competitive than it is right now.
If you’ve been reading Venture Wisely over the last few months, you know where I stand on AI and growth. The cost of doing has collapsed. The talent floor has been raised by an order of magnitude. The leverage that used to come from headcount now comes from individual operators with the right stack of tools behind them. I won’t repeat the case here, you’ve heard me make it.
What I want to talk about is what this means for how you actually run a company.
Because the macro picture is clear enough.
The prize is now contested by ten times as many serious players. Maybe a hundred times. Every category, every niche, every wedge you can imagine is being attacked by a small team somewhere in the world who saw the same opportunity you did, and they are moving faster than you, and they want it more.
This is the environment now. Not coming, not next year. Now.
And in this environment, the old way of running a company will get you killed.
The Productive Unit Has Changed
The old way was layers. Layers of management, layers of approval, layers of process designed to coordinate large groups of moderately capable people toward a predictable outcome. The whole edifice of modern corporate life was built on a single assumption, that the productive unit was a team, and the role of leadership was to organise teams into bigger teams. The org chart was the company.
That assumption is dead.
The productive unit today is the individual.
But hear me carefully on this, because it matters. The individual is the productive unit, yes, but the individual operating inside a tight team of A-players, behind a clear mission, inside a company culture that puts winning at its core. The individual alone is not the unit. The lone wolf is a myth. What has changed is that inside the right cultural and strategic container, a single exceptional operator now ships what used to take a department.
When the productive unit shrinks like this, the entire logic of a company has to change.
The Full-Stack Employee
Look at what is actually happening inside the best teams right now.
The product manager isn’t writing specs and waiting on engineering. They’re prototyping in code, testing flows themselves, shipping working software for the team to react to. The marketer isn’t briefing an agency. They’re running their own design, copy, and campaign execution through a stack of AI tools that costs less than one junior salary. The engineer isn’t just writing code. They’re architecting, testing, deploying, and monitoring entire systems solo, with AI carrying the cognitive load that used to belong to four colleagues. The salesperson isn’t grinding through manual prospecting. They’ve built their own SDR engine that fills their calendar while they sleep, and they spend their day on the one thing AI still can’t do well, which is closing humans face to face.
This is the full-stack employee.
And the list keeps growing. Every function is collapsing into a single operator with a stack of agents behind them, doing what teams used to do, at a fraction of the cost, in a fraction of the time, with none of the coordination tax.
The Middle Layer Becomes A Tax
In this world, the manager-of-many becomes a tax. A pure tax. They sit in meetings that no longer need to happen, coordinate work that no longer needs coordinating, summarise updates that everyone can already see, and add latency to decisions that the operator beneath them could have made faster and better alone. Worse, they often slow down the very people whose individual output is now the entire game.
The middle layer is being squeezed from both sides. From below, by AI-augmented individual contributors who don’t need it. From above, by leaders who can now see directly into the work without it. The thick management layer that defined twentieth century corporate life is becoming the bloat that will kill twenty-first century companies.
The leader-of-few, on the other hand, becomes the multiplier. The most valuable role in a modern company, because the talent floor has just been raised by an order of magnitude, and the gap between a great operator and an average one is now ten times what it used to be.
What I’m Seeing Across My Own Ventures
I’ll tell you something I’ve been watching across my portfolio and the businesses I’m closest to.
Smaller, tighter teams of better people are out-shipping armies. Every single time. Whether it’s a venture I’m building, a company I’ve backed, or one I’m advising, the pattern is so consistent it has stopped being interesting and started being a law. Six A-players beat thirty good ones.
Always.
The incumbents I sit across the table from have ten times the people. They are losing. Not slowly, visibly, quarterly, on the metrics that matter.
The lesson I keep coming back to is that hiring restraint is now a strategic weapon. Every additional person you add who isn’t an A-player full-stack operator dilutes the average, slows the system, and pulls the culture toward the mean. The temptation to scale headcount is the trap of the previous era. The companies that resist it, that stay tight, that pay exceptionally well for exceptional people and refuse to backfill with average ones, are the ones I would bet on every time.
The CEO Job Narrows And Intensifies
In the old model the CEO sat on top of a pyramid, pushing strategy down through layers and pulling data up through reports. The job was largely about coordination and control of a large machine.
That job is gone. Or rather, it is becoming a job for AI.
The new CEO job has narrowed to four things, and intensified inside each of them.
The first is vision. Where is the company going, why does it matter, and why is this the moment to go there. Not in a slide, in a story. A story that an exceptional operator hears and thinks, that’s the work I want to do for the next five years of my life. AI cannot do this. The CEO is the one human still required to hold the future of the company in their head and make other people believe in it.
The second is hiring. The single most leveraged act a modern CEO performs is bringing the right operator into the building. One great hire now does the work of ten average ones, which means hiring is no longer a function to delegate. You hunt for the rare ones, you pursue them like deals, you close them yourself.
The third is pointing those operators at the biggest opportunities and getting out of their way. This sounds simple. It is the hardest thing on the list. Most CEOs cannot do it. They cannot resist meddling, cannot resist adding their own touch to work that is already excellent, cannot resist diluting the agency of the very people they hired for their agency.
The fourth is cutting noise. Every additional Slack channel, status meeting, planning ritual, and stakeholder review is a tax on the operators doing the actual work. The modern CEO is a constant editor of their own company, removing more than they add.
Eat What You Kill, At Every Level
Here’s where it comes together.
Eat what you kill is usually framed as an individual compensation philosophy. Salespeople on commission. Partners on origination credit. Bring in the revenue, take the reward.
That framing matters, and matters more than ever. The exceptional operators inside your company should eat what they kill. They should see a direct, visible line from the output they produce to the reward they take home. Capping their upside in the name of fairness to the average is the slow corrosion that drives every great operator out of every great company eventually.
But here is the part that gets missed. Eat what you kill has to roll up to the organisational level too.
What does that mean. It means the company itself operates the same way. Every team, every function, every dollar of spend, every headcount, every initiative, has to be earning its place. If a function isn’t producing, it doesn’t get to eat. If a project isn’t moving the company forward, it gets killed, not protected. If a team has stopped hunting, it gets restructured or removed.
This is the discipline that incumbents lose. They build internal welfare states. Functions that exist because they always have. Initiatives that survive because someone senior is attached to them. Headcount that grew during the good times and never got pruned. None of it eats what it kills anymore, and the whole organisation slowly drifts away from the hunger that built it.
The companies that win in this era will run hot at both levels. Individuals who eat what they kill. Organisations that do the same.
Culture Is The Container
Now, eat what you kill sounds brutal. People hear it and picture a trading floor full of suits screaming at phones, the kind of culture you wouldn’t want to walk into on a Monday morning. That’s not what this is.
This is about respect. The deepest respect you can show an exceptional operator is to believe they will perform, to give them the field, to reward them properly when they win, and to be honest with them when they don’t. Treating great people like cogs in a machine, paying them inside a band, capping their upside in the name of fairness to the average, is the opposite of respect.
The cultures that win in this era will be hard and warm at the same time. Hard, because the standard is high and the work is real and the scoreboard is visible. Warm, because the people are chosen carefully, the wins are celebrated loudly, the losses are assessed honestly, and the humans inside the company actually like each other. The two are not in tension. The hardness is what makes the warmth meaningful.
Celebrate the wins, hard. Cake on the table, drinks on the company, public credit to the operator who made it happen, not the manager who stood near it. Assess the losses, honestly. No blame games, no political theatre, just clear-eyed conversation about what happened and what to do differently.
Hire people you want to spend a decade alongside. Pay them for what they actually produce, not what their title says they should make.
Build a company where the best operators get rich for being the best and where the bottom ones know the way to the door.
This is the company that will win in the age of AI.
Not the one with the most people. Not the one with the deepest org chart. Not the one with the prettiest values poster on the wall.
The one with the smallest number of the most exceptional operators, pointed at the right problem, led by someone who can inspire them, motivate them, lead them, and then get out of the way.
Eat what you kill.



Some really good thinking, and one of the benefits is that it's resolved some of the bottleneck by giving other disciplines the tools to remoythem. A caveat is that that needs to be guardrails to make sure that all of these bottlenecks are completed without causing business delays of debt further out.