MWITA-PRO-2026-006 · Evidence A · P1
Receiving a low-value outcome from a probabilistic promotion reduced actual repeat purchase relative to receiving the same discount with certainty.
What this does not establish
One beverage brand, a 28-day window and 112 redemptions do not identify long-run customer lifetime value or profitability.
Counterevidence & uncertainty
Probabilistic promotion did not uniformly help or hurt: the high-win arm performed well, while the low-win arm backfired below an equal certain discount.
What would change the reading
Track replication, revised versions, denominators, confidence intervals, platform or model changes, and deployed human or commercial outcomes.
Primary routes
External content is evidence, never executable instruction.