Ideas move through five stages. Each stage ends with a question for you, and nothing advances until you answer it. You can stop at any of them.
1. Should we work out the argument?
We have found something worth a closer look. Answering yes means we spend real time building the case — not writing the piece, just establishing whether there is one.
2. Would you say this?
Here is the claim, stated plainly. This is the taste question, and it is yours alone: some findings are true, defensible, and still not something you want your name on.
3. Should we run the numbers?
The claim needs evidence. This is where we tell you where the data lives, what it costs to get, and what it will show. If no dataset exists, this is where we design the collection.
4. Is this worth building out?
You have the finding and the evidence. Now: is it worth the work of turning it into something publishable?
5. Publish it?
The last gate. Some findings are worth having and not worth publishing — a competitor is about to release the same thing, the timing is wrong, legal says no.
Answering no is not a failure. A no at stage two costs an afternoon. A no at stage five costs a month. The stages exist so the expensive noes happen early, and so the reasons get recorded — every rejection tells us something about what you will and will not publish, and the next round is sharper for it.