Observing how similar AI investment cases fare across different organizations reveals a factor that matters as much as the specific numbers in the business case: the existing level of trust between the finance function evaluating the investment and the team proposing it, built up over prior, unrelated decisions.
Why trust affects the outcome independent of the numbers
A team with a track record of honest, well-calibrated projections in past initiatives gets more benefit of the doubt on a new AI investment's optimistic assumptions than a team whose past projections have proven unreliable, even when the current proposal's underlying numbers are structurally similar. This isn't irrational - it's a reasonable Bayesian update based on real track record, but it means the same proposal can succeed or fail based on history unrelated to its own specific merits.
Why this matters for how proposals get built
A team with less accumulated trust benefits from building unusually conservative, well-documented assumptions into an AI investment case, specifically to overcome a skepticism the numbers alone might not otherwise clear. A team with more accumulated trust has more room to make a case with less exhaustive documentation.
Why this dynamic is rarely discussed openly
Acknowledging that a proposal's actual approval odds depend partly on organizational history rather than the current numbers alone is an uncomfortable thing to say explicitly in a finance meeting, even though everyone involved is implicitly aware of it.
What this means practically for teams proposing AI investment
Understanding your team's actual standing with finance, honestly, and calibrating how much independent documentation and conservative framing a specific proposal needs accordingly, is a practical skill that matters as much as the underlying business case itself.