Here’s Actually How to Hit Your KPIs

You’re asking your team to solve everything themselves, and then you’re complaining when it’s hard to find people capable of handling that much complexity.

Your solution to scaling faster than ever before: Set aggressive KPIs. They’ll figure it out.

Your solution to implementing revolutionary AI technology: Set aggressive KPIs. They’ll figure it out.

I’m noticing a pattern. Leadership by KPI.

Im not saying setting goals is not important, but these days it seems like the data is outweighing direction.

It’s important to set the pace, but then you need to strategically plan out how to get there.

Or else you get risk, burnout, chaos, and a company that still calls themselves a start-up 20 years in.

The complexity your people are drowning in isn't a talent shortage. Your structure produced it. Or rather, your lack of structure did.

Conway’s Law teaches us that the systems we design mirror our organization’s communication structure.

What you produce as a company is not defined by how hard your people work. It’s defined by the structure you set around them: how people operate, communicate, make decisions, and execute.

So if the work is impossibly complex to execute, that complexity came from you, not from the market for project managers.

You can’t hire overqualified people to compensate for a lack of thoughtful design. You actually compound the chaos when you do.

Every intelligent expert wants to act on their ideas. So now you have ten smart people independently trying to fix the same broken system, creating ten conflicting efforts playing out in silos.

So now you have 10 conflicting efforts playing out, in silos, inharmoniously.

This is why setting KPIs alone doesn’t work. You set the target, you communicated the target, and then you left every person to independently invent how to hit it. And they each independently drew up a plan.

You need to create a system to foster a more collaborative culture - for which to channel everyone’s discussions, ideas, and efforts productively.

Atomic Habits explains this beautifully in a more personal setting. But it’s a great example of the same lesson: You don’t rise to the level of your goals, you fall to the level of your systems.

You said “Lose 10lbs” but the scale didn’t move until you placed your tennis shoes by the bed.

Until you invested in the nicer gym.

Until you put a post-it note on your fridge telling you to not eat past 9pm.

You built a system that supported the goal.

“Hit the number” is willpower as a management strategy, and it fails in a company for the same reason it fails in a kitchen at 11pm. We’re human.

As humans we always do what is easiest in the moment. What’s familiar. To operate at a new pace, or using new tools, it requires conscious energy.

And when your team is running a mile a minute, they default to familiarity because they only have so much conscious energy to spend.

If a system isn’t in place, they forget to dot the i’s. They guess. They take a SWAG.

You never gave them a playbook, so how can you be surprised when execution doesn't go smoothly?

Stop asking for more effort. Design the system so that the right moves are stupid simple & easy to execute on.

1. Define what you want.

Ben Horowitz tells a story in The Hard Thing About Hard Things about being frustrated with how his product managers were operating. Then he realized the problem: he had never actually told them what good looked like.

So he wrote it down. Good Product Manager / Bad Product Manager was essentially a one-page definition of what his basic expectations were of a PM. We are talking how to hold a meeting, send follow-up notes, etc.

Once the expectation was explicit, people were easily able to perform accordingly.

This sounds obvious, but companies skip it constantly. We set a KPI, assign an owner, and assume everyone shares the same definition of what good execution looks like.

Before you ask people to perform better, define what better actually means.

2. Get it out of people’s heads.

David Allen, author of Getting Things Done, says it perfectly: “Your mind is for having ideas, not for holding them. Get it out of your head into a trusted system.”

For companies, this means recognizing when the process is sitting in your best PM's head.

They know who needs to be involved, what questions to ask, which risks matter, when something needs to escalate, and how to get a project unstuck.

Put it on paper and free up some of their headspace for new problems to solve.

Turn the things they do repeatedly into a process someone else can follow. You’re not trying to make your experts less valuable. You’re trying to stop using their brainpower to solve the same problems over and over again.

3. Design the path of least resistance.

Once you know what you want and have the process out of people’s heads, make it stupid simple and easy to execute on.

Tiago Forte’s PARA framework separates Projects, Areas, Resources, and Archives. The organizational equivalent matters: a project has an end date; a portfolio doesn’t. Push both through the same tracker and then wonder why nobody can find anything.

Easy isn’t just a nicer interface. Easy means knowing where something lives, what you’re supposed to do, and what happens next.

The same principle applies to priorities.

Derek Sivers writes about publicly declaring his current focus as a way of escaping optionality. Once you’ve decided what matters, everything else becomes easier to reject. Clarity comes from constraint.

Organizations need the same thing.

There are a hundred ways someone could get the job done. A good operating system intentionally removes ninety of them. Not because you want robots, but because every unnecessary decision takes attention away from the decisions that actually require judgment.

Define where information lives. Define the lanes. Define the escalation paths. Define what doesn’t matter.

Then build the process into the reporting itself, so reading what’s expected and reporting status are one action instead of two.

The goal isn’t to control how people think. It’s to eliminate the things they shouldn’t have to think about.

4. Automate it.

And then AI gets interesting.

Once the process is defined, documented, organized, and constrained, you’re less reliant on experts to execute every part of it. Someone competent can get the job done well.

And AI is becoming quite competent.

Layering AI on top of an organized system unlocks a major cheat code to productivity. Soon your staff can handle more projects. Reporting gets lighter. Knowledge becomes reusable. Capacity expands without requiring headcount to expand at the same rate.

But the system comes first, not the aggressive KPI. Build it into the system so the expansion is seamless.

We’re asking people to deliver more than ever in less time than ever, and now we’re handing them AI on top of it and saying, go.

If you want people to do more, stop asking for more effort.

Define what you want. Get it out of people’s heads. Make the right moves stupid simple and easy to execute on. Then automate it.

Your aggressive KPIs can’t be met by work ethic alone. They require a system designed to make that level of performance possible.

Set the pace. Set the target.

Then design the company that can actually hit it.