Knowing when an inquiry is actually finished

Reporters rarely run out of things to check. There's almost always one more call to make, one more document to request, one more angle worth a day of your time. Left unmanaged, this tendency means an inquiry can continue indefinitely, and the decision to stop reporting and start writing ends up feeling arbitrary — dictated by a deadline rather than by any real sense that the work is done.

The alternative isn't to report less thoroughly. It's to separate two different questions that tend to get merged into one. The first is "have I answered everything I could theoretically ask?" — which is almost never fully true and shouldn't be the bar. The second is "have I answered, or explicitly set aside with reasoning, everything the published account depends on?" — which is answerable, and is the actual test that matters before publication.

The difference between open and unresolved

Not every loose thread needs to be tied off before publication. Some questions turn out to be interesting but not load-bearing — they don't change whether the central claims hold up. Others are genuinely unresolved in a way that matters: a figure that couldn't be independently confirmed, a party who was never reached for comment, an alternative explanation that hasn't been properly ruled out. The completion question isn't whether any threads remain open. It's whether the ones that matter have each been either resolved or consciously, defensibly set aside.

• List every claim the published account depends on, not every question ever asked
• For each one, mark it answered, set aside with a stated reason, or genuinely unresolved
• Treat "genuinely unresolved" items as a decision point, not something to quietly drop

"Finished doesn't mean every question is answered. It means every remaining question has been looked at on purpose."

Approaching everyone concerned properly

A completion check also has to confirm that everyone whose conduct or reputation is affected by the story has had a fair, proper opportunity to respond — not as a formality, but because it often changes what "finished" actually looks like. A response that arrives late in the process can close one question and open another, and a completion review is the natural point to confirm those opportunities were genuinely extended, not just attempted and abandoned.

None of this needs to happen alone. A second, less invested reader — someone who hasn't lived inside the material for months — is often better positioned to spot the difference between a thread that's genuinely closed and one that's simply gone quiet from fatigue.