Decide what to build first, with the reasoning written down.
RICE, MoSCoW, value against effort — the framework matters less than being explicit about the inputs. I make the assumptions visible, so a decision can be revisited when they change instead of relitigated from scratch.
