Nerve Center

Handing off a decision and handing off the reasoning behind it are two different transfers. The second one is usually the more expensive of the two, and it's the one that quietly doesn't happen.

September 10, 2026 7 min read

A product manager says she can't hand a pricing decision to someone else on the team, because no one else has the full context. It sounds like she's protecting her own turf. Look closer and the actual obstacle is more mundane: the context she'd need to hand over lives in a dozen scattered conversations from the past two weeks, half of them not written down anywhere, and reconstructing it for someone else would take about as long as just making the call herself. The decision isn't the bottleneck. Getting someone else caught up to make it is.

Making a call and explaining how you got there are two different transfers

Handing a decision to someone else looks like a single transfer, but it's actually two. The first is the decision itself, whatever conclusion got reached. The second is everything that fed into it: which conversations mattered, which ones turned out to be noise, what got said in a hallway that never made it into a doc, why one signal outweighed three others that pointed the other way. The first transfer is usually quick. The second is where the real cost lives, because most of that material was never written down anywhere in the first place, it exists only as something one person happened to notice and remember.

That's a different problem than reluctance to delegate. A person can want, genuinely, to hand something off and still not do it, because the honest cost of transferring the reasoning is higher than the cost of just deciding themselves. Every time that trade gets made, the decision stays with the one person who happened to accumulate the relevant scraps, not because they insisted on keeping it, but because nobody had a cheaper way to move what they knew.

Where this gets its name

Management research documented this pattern with real field data, not just a plausible-sounding story about busy people. Henry Mintzberg first published "The Manager's Job: Folklore and Fact" in Harvard Business Review in 1975; the version linked here is the 1990 HBR Classics reprint, which carries Mintzberg's own retrospective commentary alongside the original findings. He reported on direct observation of working executives rather than surveys about how they thought their day went, and his own study of five American chief executives found they spent 78 percent of their contact time in verbal communication, not documents. That verbal information becomes the manager's real working knowledge of the organization, held in the manager's own head far more than in any file or system. Mintzberg called this position the manager as nerve center: the person a disproportionate share of an organization's live, current information actually flows through and accumulates in, by virtue of who happens to be in enough rooms and on enough calls to keep picking it up.

Mintzberg drew the direct consequence out explicitly, under the name he gave it: the dilemma of delegation. A manager can't easily hand a task to someone else, he found, not from unwillingness, but because unloading everything relevant that's accumulated in his head, what he called dumping memory, can take as long as doing the task personally. Handing off too little context and the delegate decides on incomplete information. Transferring all of it costs as much time as the manager keeping the task. Mintzberg observed real managers caught between those two bad options, not choosing badly, just facing a real cost with no cheap way around it, decades before anything existed that could have closed the gap.

The honest limit: writing more things down doesn't automatically fix this

It's tempting to hear "the information isn't written down" and conclude the fix is a policy: write more things down. That's not quite what the actual bottleneck is. Plenty of organizations already have wikis, decision logs, and meeting notes, and the dilemma persists inside them anyway, because the hard part was never storage. It's synthesis: which scattered fact from which conversation actually mattered to which decision, and why, is a judgment call sitting on top of the raw record, and a pile of unsynthesized notes doesn't do that judgment for anyone. A wiki page that says a customer mentioned pricing concerns on a call doesn't tell the next person whether that comment was the actual reason a deal slipped or an offhand remark that meant nothing.

This also isn't a claim that every decision should be delegated, or that keeping some judgment calls with the person who has the deepest context is always wrong. Some decisions genuinely need the person holding the most context to make them. What Mintzberg's finding rules out is the more common failure: a decision NOT getting handed off, and staying stuck with one overloaded person, for a reason that has nothing to do with whether someone else could have made a good call, and everything to do with the transfer cost of the reasoning behind it.

Where this fits at Brief

This is the exact gap a product context layer is built to close: not adding another place to write things down, but capturing the reasoning as it forms, so it's structured and attributable to a decision rather than sitting as a fact that only means something to the one person who remembers the meeting it came from. That doesn't make every decision delegable. It changes the actual arithmetic Mintzberg described: reconstructing the reasoning behind a call stops being a fresh, costly reassembly job every time, because the connective tissue between the signal and the decision was captured once, when it was actually formed.

Think of the last decision you didn't hand off, even though you wanted to. Was the real obstacle that nobody else could have made the call, or that getting them the reasoning behind it would have cost you as much time as making it yourself?

Frequently asked questions

What does "nerve center" mean in Mintzberg's research? The position a manager occupies by virtue of the volume of live, informal information, much of it verbal and never written down, that flows through them because of who they talk to and what meetings they're in. Mintzberg found this made the manager's own head the organization's actual store of current information more than any formal system.

What is the "dilemma of delegation"? Mintzberg's term for a specific bind managers face, caught between two costly options: skip the context and risk a worse call from whoever they hand it to, or take the time to "dump memory," transferring everything relevant, which he found could eat up as much time as the task itself. Neither side of that trade is free, which is why a decision often just stays with whoever already has the full picture.

Doesn't writing more things down solve this? No, most teams already have notes and wikis, and the same bind shows up inside them anyway. A note recording that a fact was mentioned doesn't record why it mattered or what it changed, and that missing piece is exactly the part someone else would need to pick the decision up.

Does this mean every decision should be handed off to someone else? No. Plenty of calls genuinely belong with whoever holds the deepest context. What this points at is narrower: decisions that were fine to delegate stay put anyway, not because anyone judged the delegate incapable, but because moving the reasoning across turned out to cost as much as just keeping the decision.