The decision note records three things, and the third is the one that matters later: what you chose, what the runner-up was and what separated them, and what you knowingly accepted.
Every choice has a known weakness at signing. Six months on it is causing pain and the conversation becomes "why did we buy this" — as though nobody noticed. Written down, the same conversation is "we knew, we accepted it because of X, is X still true?" That is a decision being revisited. The other one is just blame.
But the failure this slot most needs to prevent is quieter than a bad choice. Shelfware persists because of auto-renewals and a lack of centralised visibility — that is the mechanism the utilisation data points at, and it is not a decision anybody makes. It is a decision nobody makes. The careful evaluation happened; the renewal did not get one, because by then the reasoning had evaporated and no single person owned the question.
So write two things down at signing, on the day you are still thinking clearly:
- What would make you reverse this, concretely enough to notice: a price change past a threshold, a capability the runner-up ships, a volume you outgrow, an acquisition.
- The renewal and notice dates, somewhere a human will see them at least a quarter before the window closes.
The most common outcome of a careful evaluation is an automatic renewal nobody chose.