Sometimes the smartest thing a leader can do is lose control on purpose

There's a version of good management that says a leader's job is to keep things orderly: clear direction, consistent decisions, no surprises. Andrew Grove argues that during the messy middle of a strategic inflection point, that instinct is close to backwards.

Grove describes the confusion that grips a company once everyone senses that something fundamental is changing but nobody can yet say what to do about it. His advice for this phase is blunt: let chaos reign. Loosen the level of control the organization is used to. Let people try different products, different sales channels, different customers, things that wouldn't normally be tolerated, because only stepping outside the old ruts produces the insight nobody can plan their way to.

Intel lived this literally in the late 1980s, when a debate over two competing ways of designing computer chips, called RISC and CISC, threatened to tear its microprocessor business apart. Some engineers believed the newer RISC approach was so much faster that it represented a 10X threat to everything Intel had built its business on. To get around a company culture built on compatibility, they quietly developed their RISC chip as a supposed side project, hoping its power would eventually speak for itself. It did well enough that Intel ended up introducing two major, competing chips at almost the same moment, splitting resources, confusing customers, and dividing the very engineers who had to support both.

Grove let that chaos run for a while, because it wasn't obviously wrong to explore it. But he is equally clear that the exploring phase has to end. Once the market's reaction to Intel's compatible chip made the answer obvious, Grove reined the chaos back in and the company refocused entirely on the direction that had won.

What makes an organization survive this whiplash, Grove suggests, is a kind of balance between two very different sources of insight: middle managers close to the daily numbers and customers, who sense change first but can only act locally, and senior leaders, who see the whole company but are often the last to feel the wind shift. When both are strong and both take turns leading, first loosening things to explore, then tightening things to commit, an organization can find its way through. Companies that lacked this balance struggled badly: Apple's own chief executive later admitted, in public, that he understood what needed to change but wasn't able to move the company past its old habits in time.

When you sense something is shifting but you're not sure what to do about it yet, resist the instinct to tighten the rules and instead give people room to try genuinely different approaches, even ones that wouldn't normally be tolerated. Make sure the debate that follows includes the people closest to the actual work, not just the most senior voices in the room, because uncertainty is exactly when rank stops being a reliable guide to who's right. But watch for the moment the market itself gives you an answer, because that's your signal to stop exploring. Once you see it, commit fully: rein the chaos back in, put your attention and your resources behind the direction that won, and stop quietly supporting the paths you were testing, because half-committing to two directions at once will exhaust everyone faster than picking one.