Referral Reward Contribution & Redemption Calculator

Prepare a consistent cohort

Use one currency and one fixed acquisition cohort. Count qualified customers, not clicks or invitations. Contribution per customer is the average after normal variable costs and refunds over your declared horizon, before programme benefits entered separately. If the friend benefit is already deducted, enter zero for that input. Earned awards can differ from customer count because your rules can include tiers or multiple awards. Include outstanding earned awards; do not count only redemptions.

Formula and interpretation

Let N be customers, C pre-programme contribution per customer, F friend benefit cost per customer, A earned awards, R cost per redeemed award, p redemption as a fraction and K cohort programme costs. Reward cost = ARp. Total programme cost = NF + ARp + K. Contribution remaining = NC − NF − ARp − K. Full-redemption contribution replaces p with one. Cost per attributed customer divides total programme cost by N, and is undefined when N is zero.

For AR greater than zero, algebraic break-even redemption = [N(C−F)−K]/AR. A value below zero means no feasible redemption rate covers these costs; a value above one means all rates from zero to 100% fit. With AR=0 there is no redemption-dependent threshold. The tool still reports any loss from friend benefits or fixed costs. This diagnostic is not a causal ROI calculation or a valuation of accounting liabilities.

Known-answer example

N=200, C=30, F=10, A=200, R=20, p=60%, K=1,000. Base contribution is 6,000; friend cost is 2,000; expected referrer cost is 2,400. Programme cost is 5,400, or 27 per attributed customer. Remaining contribution is 600; full redemption leaves −1,000; the break-even rate is 75%. Raising redemption from 60% to 75% removes the remaining 600. These are synthetic arithmetic scenarios, not predicted redemption behaviour.

Precision, validation and limits

Counts must be whole numbers from zero to one million. Monetary inputs allow two decimals up to one million currency units; contribution alone may be negative. Redemption accepts 0–100% with two decimals. Empty values, nonfinite numbers, separators and unsupported precision are rejected. Monetary inputs are converted to integer minor units and redemption to basis points before multiplication; expected costs can be fractional minor units and are rounded only for display. No cash payout is created.

Editing clears the prior result. Copy/download retains the inputs, scope and limitations. Everything runs locally. Supply your own cost policy and review late redemptions, refunds, abuse and currency differences outside this form. An assumed redemption rate is neither a confidence interval nor evidence that unused rewards will never be redeemed. The calculator excludes any incremental contribution from the referrer's future purchase unless you explicitly model it separately.

Sources checked October 12, 2026: Shopify referral programme ideas (visible publication June 7, 2025, despite the 2026 title) and Stripe Checkout discounts (living documentation; publication date unavailable). They support reward-design options and redemption controls, not this original model or any promised performance gain.