Finding
Treatment conversion is 0.89 percentage points higher in the simulation, but modeled incremental NPV is negative at $2.20 per prospect. Neither uncertainty interval rules out zero: the lift interval is −0.19 to 1.96 points, and the conditional NPV interval is −$10.80 to $5.34.
Why it matters
The simulated result does not support scaling the incentive. Evaluate acquisition costs, losses and retention alongside conversion, then collect stronger evidence of value.
Who could use this
Consumer-finance acquisition and campaign teams deciding whether to scale an incentive.
Data and scope
A fully synthetic randomized campaign with 12,000 fictional prospects and assumed account economics.
How the analysis works
Intention-to-treat conversion comparison, segment standardization, a three-year account NPV model and a conversion-only parametric bootstrap. Eighty-one scenarios vary credit loss, funding and retention.
Limits of the evidence
All prospects and parameters are simulated. Uncertainty covers conversion conditional on the assumed economics, not full business uncertainty. This is not a lending engine or an observed customer result.