Every interesting engineering decision is a trade against something scarce. Time, people, latency budget, the patience of whoever is waiting. The decisions that go badly are almost always the ones where nobody said out loud which of those was actually biting.
Name the scarce thing first
Before anyone argues about approach, agree on what is short. A team that thinks the constraint is the calendar will land somewhere completely different from a team that thinks it is headcount, and both can be right. The argument only feels unwinnable when each side is quietly optimising for a different shortage.
Cut scope, not quality
When the constraint turns out to be time, the reflex is to keep the scope and quietly lower the bar. This is almost always the wrong way round. Scope you cut is gone and forgotten. Quality you cut is a loan, and it charges interest, and it always comes due on the worst possible morning.
Write the decision down
A decision nobody recorded gets relitigated every few months, usually by someone who was not in the room and is very confident. A short note covering what you chose, what you gave up, and what would make you revisit it takes ten minutes and saves you that argument three times over.