By the time information reaches you, it has passed through people. Each of them summarized it, and each summary dropped the specific, awkward, telling detail in favor of the general and the presentable. Some of them had opinions about what you should conclude, and shaped what you received accordingly. None of this required anyone to lie.
You end up with a picture that is accurate in outline and useless in the particular — which is exactly the picture General Hamilton had of Suvla Bay while twenty thousand of his men sat on a beach four miles from winning the campaign.
The military historian Martin van Creveld named the instrument that fixes this, and Robert Greene borrows the term: a directed telescope.1
Van Creveld's metaphor is precise and worth taking literally. The regular reporting system — the formal chain, the standard summaries, the scheduled updates — tells a commander which questions to ask. The directed telescope is what lets him answer them.1
It is a separate, informal channel: people scattered through the organization and outside it who can give you instant, unfiltered information from wherever you point them. Not a replacement for the chain of command, which you still need. A supplement that runs alongside it and can be aimed.
The two systems work together. Formal reporting produces structured but degraded information at predictable intervals. The telescope produces unstructured, specific information on demand, tailored to whatever you need to know right now.
Greene's central case is Napoleon, whom he calls the master of this game.1
Napoleon built what amounted to a shadow organization: a group of younger officers positioned across every area of the military, selected specifically for loyalty, energy and intelligence. They held no special rank and commanded nothing.
At a moment's notice he would send one of them to a distant front, or a garrison, or — with a diplomatic cover story — into enemy headquarters itself, carrying secret instructions to gather exactly the information he couldn't get fast enough through normal channels.1
The value was in the aiming. A formal report tells you what someone thought was worth reporting last Tuesday. A young officer sent to a specific place with a specific question comes back with the answer to the question you actually have, today.
George Marshall ran a quieter variant of the same instrument, and the contrast is instructive because his organization was enormous and domestic rather than a field army.
He would sometimes drop in on an army base incognito, to see with his own eyes how his reforms were actually landing rather than how they were being reported.1 He also read letters from ordinary soldiers — a channel that bypassed every officer between him and the men, and that nobody in the chain could shape.
Both moves solve the same problem Hamilton had: an executive whose only access to reality runs through people with reasons to manage what he sees.
Greene is honest about the limitation, and it's the reason the telescope has to be directed rather than simply comprehensive.
Marshall's incognito visits worked, but in an organization of any real complexity, going to look yourself consumes far too much of your time to be the general answer.1 You cannot personally verify everything, and a leader who tries has abandoned the larger picture in order to check details — the exact failure Marshall's whole approach was designed to avoid.
The telescope is therefore a sampling instrument, not a surveillance system. You point it at the specific thing you currently doubt, get an unfiltered read, and put it down.
A project three levels below you is reporting green, and something about the reporting has started to feel too smooth. Every update is confident. No update has ever contained a problem you hadn't already heard about from somewhere else.
The wrong move is to demand more detailed reports, which just produces longer documents written by the same people with the same incentives. The other wrong move is to go inspect it yourself, which burns your time and signals distrust loudly enough to change the thing you're trying to observe.
The telescope move is narrower. Find one person close to the actual work whose judgment you have independent reason to trust — someone who is not in the reporting line for this project and has no stake in how it looks. Ask them one specific question. Not "how's it going," which invites a summary, but something concrete enough that the answer can't be smoothed: what's the thing on this project that people are hoping nobody asks about?
Then put the telescope down. The point is a single unfiltered read on a specific doubt, not a parallel intelligence service that makes your own chain of command redundant.
The concept comes from van Creveld's Command in War (1985), a serious work of military history, and Greene is transmitting rather than inventing it. The Napoleon and Marshall cases are both well documented. [POPULAR SOURCE] for Greene's framing.
Open tension the chapter leaves entirely alone: a network of loyal informants who bypass the formal chain is, from the perspective of everyone in that chain, a surveillance system reporting on them to the boss. Greene presents it purely as an information-quality tool and never addresses what it does to the people being observed.
Napoleon's officers knew the shadow brigade existed. That knowledge is itself a management instrument — and a different one from the stated purpose, closer to the effect of Marshall's $3.52 grass-seed remark than to neutral fact-gathering. Whether the telescope's real function is better information or the fear of better information is a question the chapter never asks, and the answer probably differs case by case.
Set this against Strategy 5: The Command-and-Control Strategy, the chapter this concept sits inside. That chapter's central instruction is to build and protect a clean chain of command — unity of command, clear orders, no ambiguity about who decides.
The directed telescope is an explicit admission that the chain will never be good enough. However well you build it, information will degrade passing through it, and you will need a way around your own architecture.
The pairing produces something the chapter doesn't state: a chain of command is not a truth-transmission device and was never designed as one. It's a decision-transmission device, optimized for getting instructions down rather than reality up. The two functions run in opposite directions through the same structure, and the telescope exists because the structure can only be optimized for one of them at a time.
Business → Skip-Level Meetings and Their Predictable Corruption. Most large companies formalize some version of this — skip-levels, open-door policies, anonymous surveys, engineering "office hours" with executives. The insight the pairing produces: formalizing a directed telescope tends to destroy it. The moment the channel becomes official and scheduled, it acquires its own incentives, and people prepare for a skip-level exactly the way they prepare a status report. Napoleon's version worked partly because it was irregular and unannounced. A telescope on a calendar is just another reporting line with a friendlier name.
Psychology → Information Loss in Serial Reproduction. Bartlett's classic serial-reproduction research demonstrated that a story passed through a chain of people doesn't just lose detail — it becomes systematically distorted toward the conventional and expected, with the anomalous elements dropping out first.
The insight: the specific things a leader most needs from the field are precisely the ones the chain is most likely to delete. Anomalies are the early signal of everything important, and they are exactly what serial retelling smooths away. The telescope isn't compensating for random noise in the chain; it's compensating for a bias that runs reliably in one direction.
The higher you climb, the more people stand between you and anything real, and every one of them is being helpful. Blindness at the top is not caused by bad subordinates — it is the ordinary, predictable output of a well-functioning hierarchy, and it has to be actively counteracted or it simply accumulates.