A long article needs a question, not a container
A folder full of good material can still produce a shapeless draft. Start by deciding what the piece is trying to find out.
The notes contain an interview, three useful quotations, a set of figures and an excellent detail about a broken toaster. They do not yet contain an article.
Consider a proposed feature about a repair café that has stopped holding Saturday sessions. There is plenty to describe: the volunteers, the objects, the room, the ritual of tea and patient diagnosis. But “a piece about the repair café” gives the material a subject without giving the writing a direction.
“Why did a popular session become difficult to sustain?” offers something to investigate. The answer may concern the room, the volunteers’ time, the type of repair being requested, or an assumption in the question itself. Was the session as popular as everyone remembers?
Make a provisional promise
A working question is not a conclusion you must defend. It is a way to choose which material deserves attention.
Write one sentence describing what the finished article should help the reader understand. For the repair café: “Explain what Saturday sessions demanded of the volunteers, and what changed before they stopped.” This is narrower than a history of the organisation and more useful than “explore community repair”.
The sentence can change. If the interviews reveal that Saturday sessions are merely moving to another venue, the original premise may need substantial revision. Planning should make that discovery easier to notice, rather than make the first angle expensive to abandon.
Arrange the evidence by what it can do
Chronological notes record how you encountered the material. An article often needs a different order.
The broken toaster may make a vivid opening if it reveals the patience required for a repair. Attendance figures may establish demand, but only after you understand what was counted. An interview with a volunteer may explain the workload while leaving the venue’s account unanswered.
Give each proposed section a purpose. “What a Saturday required” might need a description of preparation and follow-up work. “What changed” needs a sequence supported by dates and sources. “What happens next” should distinguish a confirmed arrangement from a hopeful proposal.
You can use headings to make that structure visible, but do not mistake the labels for completed thinking. Under each heading, put the evidence you have and the important thing you still need. Keep references beside the claims they support, so a pleasing outline does not conceal an unsupported assertion.
Give material permission to stay out
The best quotation in your notes may belong to a different article. Keeping it simply because it took effort to obtain can bend the piece away from its question.
Make a separate place for useful material you are not using. It can be a short section at the end of a working document or another file, clearly named. This is less emotionally demanding than deleting everything and less damaging than making the reader accommodate it.
Your reading and interview commentary can remain separate from source material. The plan can then link to the relevant documents without swallowing every transcript and note. When a section is ready to write, open the sources it actually needs.
Start drafting once you can see a plausible route through the question, even if some gaps remain. Mark those gaps specifically: “Confirm when the room booking changed” is useful; “more research” is an invitation to wander. The first full draft will test whether your planned sequence works in prose. Expect to move things.
Find app controls and examples in the Clarity documentation
A little more room to think.
A free, local Markdown hub for Mac.
Mac app details