AI Assistant (Leat MCP)
July 20
→
Reward and redemption analytics
Judge every reward on what it costs and what it does, not just how often it's chosen. See redemptions, repeat redemptions, cost per redemption, and what happens after someone claims one. Because cost sits in the catalog beside performance, the reward that looks popular and the reward that's actually worth keeping are easy to tell apart.
For each reward and voucher: how many members could claim it, how many did, how many came back for it again. Where members drop out between eligible and redeemed. All of it by segment, tier, location, and channel, so a reward that works for Gold in the city and dies in the suburbs is visible as exactly that. Vouchers report issued, used, and expired, so leakage and breakage are figures rather than guesses.
Cost per redemption, by reward and by mechanic. Total reward spend this period against last. Retail value against cost for product rewards, so the margin case for a free coffee over a discount is visible in the numbers. Rewards ranked by redemptions and by cost side by side, which is where the expensive favorite and the cheap unknown show up.
A redemption is an event on the customer's record, so what followed it is on the same record. Did the member return, and how soon. Did their spend change. Did a customer who received a product free go on to buy it at full price, which is the measure of a product reward as an introduction. Rewards are judged on what they did to behavior afterward, and the question of whether they caused it hands off to incrementality measurement.
Rewards nobody redeems are visible as such, and they come out. Rewards that are redeemed constantly and cost more than they return are visible too, and they get repriced or retired. The high-margin reward that's never chosen gets moved up the catalog or attached to a mechanic. And because every reward carries a track record, Decisioning has evidence when it chooses which reward to offer a customer.
Program integrity
Keep stamps, points, and rewards worth what they're meant to be worth
Margin-safe promotions
Run offers that can't combine into a loss
Seasonal and limited-time campaigns
Launch and retire rewards on schedule without manual upkeep
Product and category protection
Keep discounts off the items and categories that can't carry them
Loyalty program setup
Define how the program works before the first customer joins
