The choice is not between hands-on or hands-off. In our research, the entrepreneurs who led their companies from start-ups into some of the greatest corporations in history generally had both a hands-on style and an empowering style. No matter how big their companies became, they remained closely connected to their people, hyper-aware of facts on the ground, and directly engaged in strategic imperatives. If you lose your voracious curiosity about tactical details, if you lose passionate interest in people and how they are feeling, if you insulate yourself in the protective cocoon of executive comforts, you may well wake up one day to discover your company has already entered a doom loop of decline and self-destruction.

People have this vision of what it’s like to be an executive or CEO or leader of a huge business unit. They assume everyone at that level has enough experience and savvy to at least appear to know what they’re doing. They assume there’s thoughtfulness and strategy and long-term thinking and reasonable deals sealed with firm handshakes.

But some days, it’s high school. Some days, it’s kindergarten.

That was true when I first joined the C-suite at Philips and when I became a VP at Apple and when I was CEO of Nest and when I entered the ranks of Google execs. All these jobs felt incredibly different, but at their core the responsibilities were the same—it was less and less about what you were making and more about who you were making it with.

As CEO, you spend almost all your time on people problems and communication. You’re trying to navigate a tangled web of professional relationships and intrigues, listen to but also ignore your board, maintain your company culture, buy companies or sell your own, keep people’s respect while continually pushing yourself and the team to build something great even though you barely have time to think about what you’re building anymore.

It’s an extremely weird job.

In a loosely coupled organization, where talent density is high and innovation is the primary goal, a traditional, control-oriented approach is not the most effective choice. Instead of seeking to minimize error through oversight or process, focus on setting clear context, building alignment of the North Star between boss and team, and giving the informed captain the freedom to decide.

CEOs often avoid these decisions because they involve executives who have become dear friends. We recognize that this is a touchy subject, but it must be faced if the organization is to grow. One option is for some of the early team members to help launch a new product or division. They are usually more comfortable in a start-up situation or working on a smaller team. And several of the early leaders might be relieved to have the burden of an increasingly important and complex function taken off their shoulders. You won’t know until you have these crucial conversations.

The lesson, and perhaps the irony, of Tsedal and Sebastian’s study is that executives who deferred to subordinates moved up the pecking order faster than those who refused to bend to their underlings’ will and wisdom: Leaders were granted more power because they gave it away. Yet deference and “flattening” the hierarchy aren’t always the right moves. The University of Michigan’s Lindy Greer shows that the best leaders are adept at “flexing” the hierarchy.