Main idea

"Run me a data analysis on this topic", "check whether it's worth it": the request has the shape of a task, but what it transmits is vagueness rather than framing. Whoever receives it has to reconstruct what the requester had in mind — why this topic is being looked at, which audience is concerned, which pain or opportunity is targeted, what value is expected, which question needs settling, which signals already exist.

Getting those elements out is not an administrative formality: it is the part of the work only the requester can do. Skipping it doesn't remove the work, it moves it onto a developer, a designer or a data analyst who will redo it less well, never having seen the original signals.

Why it matters

This separates two gestures the tool conflates: opening a topic and offloading it. Both produce the same card, with the same title, at the same place in the queue.

It also gives the limit of any intake process: what is being asked for is not a completed form, it is a state of thinking reached. A template filled in without that work passes the check and transmits the same vagueness.

Nuances and limits

There are cases where the vagueness is the topic — an exploration whose question nobody yet knows. The framing then bears on the uncertainty itself: what we don't know, and what we're trying to learn.

And the amount to write isn't fixed: it depends on what the team already shares.

Open questions

• How do you turn down an insufficiently prepared request without the refusal being read as a lack of cooperation?