Somewhere in the last decade, the document-first meeting became gospel. Write a crisp pre-read, send it ahead of time, and the meeting becomes a high-quality discussion among people who have already absorbed the context. It is a genuinely good idea, and in the companies where it works it is transformative. In most companies it does not work, and the failure has a very specific signature: the meeting opens with everyone silently reading the document that was supposed to have been read already.
You have sat in this room. The author spent the better part of a day writing a careful six-pager. Twelve people were asked to read it in advance. Two did. So the meeting begins with a fifteen-minute reading block, which means the pre-read did not save the meeting any time. It added a four-hour authoring cost on top of a meeting that is exactly as long as it would have been with no document at all.
Why the pre-read goes unread
The honest reason is that reading a pre-read is unpaid, invisible work that competes with paid, visible work. Nobody can see whether you read it. Nobody is blocked if you did not. The meeting will accommodate you either way, because it always accommodates the lowest level of preparation in the room. As long as the meeting bails out the unprepared by reading aloud or summarizing at the top, the rational move is to not prepare, and rational people respond to incentives.
This is the same dynamic that hollows out a weekly status meeting. The forum absorbs the cost of everyone's lack of prep, so the prep stops happening, so the forum has to absorb more. The pre-read does not fail because people are lazy. It fails because the structure removed every consequence of skipping it.
The pre-read is doing a job the system should do
Step back and ask what the pre-read is actually for. It exists to bring everyone to a shared, current understanding of where things stand before a decision is made. That is a reasonable need. The problem is that a document is a terrible way to meet it, because a document is a snapshot. The author froze the state of the world at the moment they wrote it, and by the time the meeting happens the state has moved, so the pre-read is both effortful to produce and already slightly stale on arrival.
When the status of work lives in a document that has to be hand-assembled before every meeting, you are paying a person to manually serialize a state that the system already holds. The four hours of authoring is four hours of a human copying numbers and statuses out of various tools and into prose, which is exactly the PMO status laundry problem wearing a more respectable outfit. The labor exists only because the underlying state is scattered and nothing maintains a live, shared view of it.
What replaces it
The version of this that works is not a better-written pre-read. It is a standing, live view of the work that is always current, so there is nothing to assemble and nothing to read in the room because everyone has been able to see the state continuously. The meeting does not open with reading because the state was never hidden inside a document in the first place. It was visible the whole time, and the meeting is free to spend its entire length on the decisions that the visible state surfaces.
This is the difference between preparing a status and having a status. When you run reviews around traceable work that is always up to date, the pre-read collapses into a link, and the link points at something live rather than a frozen six-pager. The author's four hours disappear, not because the prep stopped mattering but because the system now does the serialization that the human was doing by hand. The meeting reclaims its first fifteen minutes, and it reclaims them every single time it meets.
What to do this quarter
Take your most pre-read-dependent meeting and instrument the gap. At the start, ask honestly how many people read the document. Then time how long the meeting spends bringing everyone to a common baseline, whether through silent reading or a verbal recap. That number is the cost the pre-read was supposed to eliminate and did not.
Then ask the harder question: how much of the pre-read is the author transcribing state that already exists in your tools? Every sentence that is really just "here is the current status of X" is a sentence the system should be able to render live. The pre-read worth keeping is the part that contains judgment, argument, and a recommendation. The part that is a hand-typed status report is the part a live view should have killed.
FAQ
Why doesn't anyone read the pre-read? Because reading it is invisible, unpaid work and skipping it carries no consequence. The meeting always accommodates the least-prepared person by reading aloud or recapping, which makes not preparing the rational choice. The structure removed every penalty for skipping, so people skip.
Aren't document-first meetings supposed to be more effective? They are, in companies where reading actually happens and the document holds judgment rather than status. The failure mode is using the pre-read to hand-assemble a status report, which is stale on arrival and gets re-read in the room anyway. Keep the judgment, kill the transcribed status.
What should replace the status portion of a pre-read? A live, standing view of the work that is always current, so there is nothing to assemble and nothing to read in the room. When status is traceable and always up to date, the pre-read shrinks to a link plus the author's actual recommendation.
Doesn't writing the pre-read have value even if unread? Writing clarifies the author's thinking, which is real. But four hours of transcribing state that already lives in your tools is not clarity, it is manual serialization, the same labor behind PMO status laundry. Keep the thinking, automate the status.