Founder mode, for design leaders

Paul Graham’s Founder Mode, written after a talk Brian Chesky gave at Y Combinator, went up on September 1. Its argument: the standard advice for scaling a company, hire good people and give them room, quietly turns founders into managers who no longer know what’s happening in their own product. Within a week “manager mode” was an insult.

It landed hard on design leaders, after two years of flattening. Meta had made 2023 its year of efficiency, and the message from the top of a lot of companies was the same: fewer layers, more people on the tools, leaders closer to the work.

Know the work

Good design leadership was always in founder mode about the work. It looks at screens, not status. It knows which flow is weak and why, and can point at the state that’s wrong without anyone briefing it first.

That’s the useful core of the essay for design: you can’t hold a quality bar from a distance. A design leader who only sees the work in a roadmap deck is reviewing the plan, and the plan was never where quality went wrong.

Try this: could you name, right now, the weakest screen your team shipped this quarter, and say what’s wrong with it? If the answer comes from a status report, you’re in manager mode.

Don’t become the queue

The dangerous version is mistaking presence for control. A leader in every review, approving every decision, becomes the queue. The team stops making calls, because the call will be remade upstairs anyway. Quality goes up for a quarter, and judgment across the team goes down for a year.

At Uber my title was practice lead: a senior designer doing the work who also managed designers. That put both halves of this in the same week, and the line I learned to draw was between the surfaces I owned and the ones I only reviewed. Go deep on the few surfaces that set the bar, and say out loud why. Then make the bar legible enough that people can hold it without you: a written standard, and critique with reasons attached. Being in the details is how you teach them.

Less in the way

If leaders are going to be closer to the work, there has to be less between them and it: fewer status meetings, fewer layers between the person drawing and the person deciding. Companies were already doing that before the essay gave it a name, and it has cost people their jobs. Flattening only helps when the people who remain share a standard they can apply without a manager in the room.