The cost of a document you generated is paid by whoever has to read it

Production got cheaper and reading did not

The effort of writing used to act as a filter. A four-page report cost somebody an afternoon, and that cost quietly prevented a certain amount of unnecessary writing, because nobody spends an afternoon on a document they are not sure anybody wants.

That filter has weakened considerably. The same four pages can now be produced in the time it takes to describe them, and the natural consequence is more documents, longer documents, and documents produced speculatively in case they are useful.

Reading, meanwhile, costs exactly what it always did. The asymmetry is the whole story, and it is felt not by the author but by everybody downstream.

It is worth noticing that this is an externality in the ordinary sense. The person deciding whether to produce the document does not bear the cost of it being produced, so the decision is made on the wrong side of the ledger, and no individual is behaving unreasonably at any point in the sequence.

The characteristic output is longer than it needs to be

Generated material tends towards completeness. It covers the obvious objections, restates the background, includes the section that any document of this type would include, and arrives with the shape of a thorough piece of work.

For the writer this feels like value, because more has been produced. For the reader it is padding, and the specific injury is that the padding is competent enough that skimming is unsafe. Obvious filler can be skipped confidently. Well-written filler has to be read to find out that it is filler.

The result is a document that takes forty minutes to read and contains ten minutes of content, and the reader has no way to know that in advance.

Structure makes this worse rather than better. Headed sections and neat lists suggest that each part carries comparable weight, so a reader working through them methodically spends the same effort on the section that mattered and the two that were included because the form seemed to call for them.

The volume lands on a small number of people

In most organisations the reading is concentrated. A handful of people read everything: the reviewer, the manager, the person who has to consolidate, the one who is copied on everything because they once asked to be. An increase in output distributes itself unevenly onto them.

The failure that follows is not that these people complain. It is that they start skimming everything, uniformly, including the documents that needed careful reading, and the degradation is invisible because nobody announces that they have stopped reading properly.

That is the real cost of the extra volume, and it is paid in the quality of attention available to the material that mattered.

The same concentration applies to approval queues, review rotas and anything else where one person is the bottleneck. Doubling what arrives at a bottleneck does not double the work done there; it changes how the work is done, usually by lowering the standard quietly enough that nobody has to admit to it.

Send less, and say what it is for

The remedies are unglamorous. Ask whether the document needs to exist before producing it, which is a question the old cost of writing used to ask on your behalf. Send the summary and keep the full version available rather than sending both. Write the request at the top: what you want the reader to do, and by when.

State the reading cost honestly. A note saying that only the second section requires attention is a genuine kindness and it is also self-interested, because a reader who trusts your signposting reads your documents properly.

Where a document was produced quickly and has not been thoroughly checked, say so rather than letting the polish imply otherwise. A reader deciding how carefully to read is making a resource decision, and they should be allowed to make it with accurate information.

Length is not the same as thoroughness, and it never was

A useful discipline is to write down what the document establishes in one sentence before producing it, and then to check afterwards whether the document says more than that sentence. Frequently it does not, and then the sentence was the document.

This applies with particular force to material that circulates: status updates, meeting notes, internal reports, summaries of things people were already at. Each of these has a natural short form that got longer once producing the long form became easy.

None of this argues against writing things down, which remains one of the more useful things an organisation can do. It argues against the specific new failure, which is producing documents nobody asked for, at a length nobody needs, and treating the resulting volume as a sign that more work is being done.