Estimate the Cost of Loyalty Points

Loyalty points create a future promotional commitment whose cost depends on the reward and its eventual use. Their stated redemption value is not automatically the same as the business cost of fulfilling a reward. Build a points rollforward and a reward-cost model before assigning a percentage of sales to the program. This article addresses internal campaign planning only; it does not prescribe accounting recognition, expiration rules, or the legal treatment of customer balances.

Translate issued points into reward units

Suppose a hypothetical program issues 100,000 points during a campaign. A merchandise reward requires 1,000 points, so those points represent 100 full reward equivalents if all are used for that reward. Assume the planning scenario expects 70% of the issued points to be redeemed in qualifying full blocks. Expected fulfillment is 70 reward equivalents. This is an aggregate expectation, not a promise that exactly 70 customers will redeem or that every account has enough points.

Use fulfillment cost rather than retail value

The reward has a stated customer value of $10, but its incremental product and delivery cost is $4 per fulfillment in this scenario. Expected fulfillment cost is 70 times $4, or $280. If campaign software and administration add $120, total modeled program cost is $400. Multiplying the $10 stated value by 70 would instead produce $700 of redeemed face value, a different measure that should not be mislabeled as the physical fulfillment cost.

A dollar-off reward needs a different model

If the same points instead buy a $10 discount on a future order, the economic effect depends on the order that would otherwise occur. On a natural purchase, the discount can remove $10 of contribution, adjusted for any avoided price-dependent fees. On a genuinely additional order, the relevant result is its net contribution after the reward. Do not automatically reuse the $4 merchandise-reward cost or assume the future purchase is entirely additional.

Track issued and redeemed balances

Starting with 20,000 outstanding points, issuing 100,000, and recording 60,000 actual redemptions leaves 60,000 points outstanding before any other explicitly supported adjustments. The redemption forecast and the actual rollforward answer different questions. Do not remove points from the customer ledger merely because the budget assumes some will remain unused. Maintain issuance and redemption cohorts so the cash forecast can place product purchases, shipping, and program invoices in realistic periods.

Test program scale and customer behavior

At full redemption of the new 100,000 points, physical reward fulfillment would cost $400 and the same administration expense would bring total modeled cost to $520. At 40% redemption, the total would be $280. These are cost sensitivities, not proof that low redemption makes the program effective. Evaluate changes in total customer contribution against a suitable no-program baseline, including concessions on ordinary purchases and any extra customer-service work.

Keep the limitations visible

Points can have minimums, alternative reward choices, or mixed redemption patterns that require a more detailed model. Use the actual program design and current supplier costs instead of a universal points-to-dollars assumption. Probabilities must lie within zero and 100%, and a zero points-per-reward requirement cannot support this division. Reserve future cash in the planning schedule where needed, but do not present the resulting allowance as a statutory liability, guaranteed savings, or expected customer lifetime value.

Method references: OpenStax: relevant information for decisions; OpenStax: financial budgets. Accessed October 22, 2026. All dollar figures are hypothetical worked examples.