Picking the Right Reward

Currency first, amount second

The currency decides who responds; the amount decides how strongly. Getting the currency wrong cannot be fixed by raising the amount — offering cash to customers who wanted status, or credit to customers about to leave, produces weak results at any price.

The five currencies, and who each one suits

Each has a different real cost and appeals to a different kind of customer. The gap between what a reward is worth to the recipient and what it costs you to give is the lever most programs underuse — and it is widest on everything except cash.

• Cash — universal, immediately understood, and the most expensive: it leaves the business entirely
• Account credit — costs near marginal cost, valuable to heavy users, worthless to someone leaving
• Discount on next purchase — cheap, but only motivates customers who intended to buy again
• Free product or upgrade — high perceived value, low marginal cost, and it deepens product usage
• Status or recognition — nearly free, and effective precisely where cash feels inappropriate

When cash is the right answer

When the purchase is infrequent or one-off, so credit has little value, and when the referrer has no ongoing relationship to deepen. High-value, low-frequency categories — financial products, home services, vehicles — are where cash consistently outperforms cheaper currencies.

When account credit wins

When customers buy repeatedly and the product has healthy margin. Credit costs you close to marginal cost, is worth full face value to the customer, and selects for people who intend to stay — which quietly improves the quality of who gets referred.

When to reward with status instead of money

Where paying a customer for a recommendation would make it feel bought — healthcare, education, professional services, non-profits. In those categories a paid endorsement damages the recommendation's credibility, which is the thing giving it value in the first place.

How to size the reward

From gross margin on the introduced customer, with headroom for the ones who do not stick. Start from margin on the first order or first contract year, decide whether the reward is one-off or recurring, and cap the total across both sides well inside that figure.

1. 01Start from gross margin, not revenueRevenue-based rewards look affordable and are not. Margin is the only figure that makes the arithmetic work.
2. 02Decide one-off or recurringA reward that repeats on renewal is a materially larger commitment than the same figure paid once.
3. 03Split across both sidesDivide a single budget between referrer and referred rather than doubling it. Two-sided splits outperform concentrating the same money on one side.
4. 04Leave headroom for churnSome referred customers will not stay. Size the reward so the program still works when a share of them leave.

Should both sides get the same thing?

Usually not. The two parties are in different positions: the referrer is committed and values credit or recognition, while the referred person has no relationship yet and needs a reason to try. Symmetry looks fair and performs worse than matching each side.

Does a bigger reward produce more referrals?

Up to a point, then it stops buying intent and starts buying volume — more signups from people who wanted the reward, fewer who wanted the product. Watch referred-customer retention when you raise a reward; if it falls, the extra spend bought nothing durable.

The mistake to avoid

Raising the reward to fix low participation. Most participation problems are visibility problems: customers have not seen the program, or were asked at the wrong moment. Increasing the reward in that situation pays more money to the few people who already knew about it.