Count what lands on your team in a week. The metrics dashboard, refreshed hourly. The automated sprint summary. The all-hands deck, the status thread, the monthly report that exists because it has always existed and the new one that exists because the tool made it easy. Each arrived with the same implicit claim: being informed is good, so more information is better.
Herbert Simon — economist, psychologist, one of the founders of artificial intelligence — dismantled that claim in a single argument in 1971, in a lecture on designing organisations for an information-rich world. What information consumes, he observed, is the attention of the people who receive it. So a wealth of information produces a scarcity of exactly the resource needed to use it. The bottleneck had moved: for most of history the hard thing was getting information; from now on, the hard thing would be affording to process what arrives.
That inversion turns every report into a purchase order. A dashboard is not a free good sitting harmlessly on a screen; it is a standing withdrawal from a fixed budget of collective attention — and the budget genuinely is fixed, because working hours and working memory do not scale with the tooling. The question Simon forces is brutally simple: does this system consume more attention than it saves?
Audit by that one question
Run the test on any information system your team maintains, and be honest about both sides of the ledger. The weekly metrics review: how many person-hours of attention does it consume, and what decisions in the last quarter actually changed because of it? The automated summary: who reads it, and what would they fail to know without it? A report that changes no decisions is not neutral. It costs its reading time, and — worse — it crowds out the thinking that has no channel of its own, because attention spent absorbing input is unavailable for the work only people can do.
Simon's design principle follows directly: in an information-rich world, systems should be built to filter and absorb, not to supply. The valuable report is the one that consumes information on your behalf — that reads a hundred signals and emits one judgment. Most organisations build the opposite: systems that multiply what must be read while leaving all the filtering to the person at the end of the pipe, then treat that person's overload as a time-management failure.
The practical move is subtraction with a measurement. Kill one recurring report or meeting for a month and watch what actually breaks. The honest answer, much of the time, is nothing — the information was being produced, not used, and its audience was attending out of politeness rather than need. Restore it only on evidence of a decision that suffered — the one review standard nobody applies to reporting.
Watch out for
Filtering has a failure mode Simon understood well from his own work on decision-making: the filter embodies a theory of what matters, and the theory can be wrong. The signal that saves you is by definition the one your current filters were not designed to pass. So subtraction needs a complement — some deliberately unfiltered contact with raw reality, customers, incidents, the floor — or the organisation becomes serenely uninformed. His own confident public predictions about how quickly machines would match human thinking were wrong by decades, which is a fair reminder that a filter built on a bad forecast filters wrongly with great efficiency.
Answer this next
Which single report or meeting consumes the most collective attention on your team — and can anyone name the last decision that changed because of it?
