Our organizations are being DDoSed from the inside

I’ve written before that you can outsource work, but not your understanding.

As a CEO, I’m finding that harder to live up to. It feels like our organizations are being DDoSed from the inside. Nobody is trying to overwhelm the company. Everyone is doing something they think is useful. But all of that work draws on the same limited capacity to understand, review, and decide. We can generate changes much faster than we can give them that attention.

I’m trying to keep product and marketing aligned while what we’re building changes daily. By the time I understand something well enough to contribute, the work might have already changed. Teams and agents are moving faster than any of us can follow.

I think there’s a pretty big problem looming below the surface in companies: leadership losing the ability to steer. Leaders never saw every piece of work. But now the work can change the direction of the company faster than we can have the ability to steer it. It’s like a car speeding up until it becomes impossible to steer.

I trust our people. We’ve put a lot of effort into hiring and building an environment where we hope people can make good decisions. Even with that effort, we’re seeing unclear or changing direction and overlapping or conflicting features. I don’t think this is a matter of people making poor decisions.

Before, we debated plans and roadmaps. There was time to understand the thinking and shape the direction. Now something can go from an idea to a working implementation before anyone outside the immediate team knows it exists. Everyone is left piecing together what happened, why, and how it fits.

When someone can build something in twenty minutes, asking to discuss it at tomorrow’s meeting starts to feel unreasonable. The thing is already there. Saying “looks good, let’s go” is easier than asking everyone to step back and consider whether it’s something we should do.

Why push back on work that’s already done and improving something?

But the fact that the work exists doesn’t tell us whether it’s the right thing to do. We have to get past treating output as proof of value. A working implementation shows that something can be built, not that it belongs in the product. Some changes have negative value.

A small change can also overlap with another team’s work or take the product in a direction we haven’t considered.

A working implementation doesn’t make those consequences disappear just because nobody thought them through.

By then, giving direction means going back to understand why a change was made, what was considered, and how it fits with everything else. This generally takes a lot longer than 20 minutes. Now this thing that someone made has made lot of people distracted from what we're supposed to make.

But moving every discussion ahead of implementation doesn’t solve it either. You can’t hold an alignment meeting for every one of the 1,000 PRs the team will open today.

So something has to change. Companies need to keep the ability to decide what work should be done, without having to discuss every change individually.

Part of that is making intent clearer ahead of time: what we’re betting on, what matters right now, what’s out of scope. People and agents can check their own work against it. It won’t catch everything. For my

The other part is being more deliberate about when an experiment becomes part of the shared product. Building something is much cheaper now. But landing it creates consequences beyond the person or team that built it.

We recently had to lock down our internal release channel because things were changing too quickly and multiple teams were overlapping. Customers weren’t seeing this, but internally we were struggling to keep up with our own changes. People can still experiment behind other feature flags, but the gate gives us one place to consider changes together before they land.

That helps us keep track of what’s changing. It doesn’t yet tell us whether all this work adds up to the direction we want to take.

And this isn’t just a problem for the CEO. I need to trust the judgement from my leaders, and they need to trust the judgement from their reports. The all the people whose judgment you rely on are also trying to keep up with everyone else.

I think this is happening at every level of the company, even if we’re reluctant to acknowledge it. It’s exciting to see how much people can do. It’s uncomfortable to admit that you’re losing the thread.

I wonder how the CEO’s role and organization needs to evolve, and the tooling needed to this. It’s tempting to step back because you don’t want to become the bottleneck or slow down experimentation.

But stepping back doesn’t remove your the responsibility for where it all leads. As CEO, I’m responsible for what the company does as a whole and whether it succeeds. Each team can make reasonable decisions about its own work, and I still need to understand what those decisions add up to, and how I provide the direction.

That’s what I’m finding hard to do. I try my best with using agents in Linear, pulling context, decisions, work across the company. Writing my thoughts on various things and sharing them. I still want our work to be good, and considered, something the customers can enjoy and we can take pride of. But my feeling is now easy to lose those parts when velocity of progress is so high.

Are we moving toward something, or just getting better at reacting? I can see more getting done. I’m less sure we share an understanding of where it’s taking us.

How are you dealing with this?