Decisions · Earn and redemption

When should you cap earning

Cap earning when a member's points would otherwise cost more than the margin their spend produces after breakage. Set the cap at the point where awarded value equals marginal contribution, usually a points-per-member limit per month or per category. Publish it before balances build; do not impose it retroactively on existing balances.

The decision

We are deciding where to stop awarding points, not whether to run a loyalty programme. The mechanism is liability recognition: each point issued is a future redemption cost that must be carried against the margin from the member's spend. If the earn rate multiplied by the redemption value exceeds the gross margin per dollar of spend, every extra transaction deepens a loss. The decision therefore turns on the crossing point between marginal liability and marginal contribution. Capping applies to continued accrual, not to points already in a member's balance.

What it costs you

A cap costs margin because you stop some high-frequency earning you could have captured. It costs member trust if announced after members have planned around uncapped balances. It costs engineering time to enforce rules at checkout or statement run, and it costs operating capacity to handle exceptions. The uncapped alternative costs more in liability, because points issued today can sit unredeemed and still represent a claim.

Illustrative calculation: a programme awards 2 points per dollar and values one point at one cent. A member spends 50,000 dollars and earns 100,000 points. The redemption liability is 100,000 points times one cent, so 1,000 dollars. If the gross margin on that member's spend is 1.5 percent, the programme earns 50,000 dollars times 0.015, so 750 dollars. Uncapped, the programme has a negative position of 250 dollars before other costs. A cap at 37,500 points holds liability at 375 dollars, leaving 375 dollars of the 750 dollars of margin. The cap converts a loss into a thin but positive contribution.

How to decide

Calculate your unit margin per dollar of member spend after payment costs, fulfilment and breakage. Multiply your earn rate by your average redemption cost per point. If the product is below unit margin, you may not need a cap on ordinary spend, though you may still cap categories with high arbitrage. If the product equals or exceeds unit margin, implement a hard cap at the member or category level where cumulative points cost equals cumulative margin. For a one point per dollar earn rate, a half cent per point redemption cost and a one cent per dollar margin, each dollar of spend creates half a cent of liability against one cent of contribution, leaving half a cent; that programme can remain uncapped on the margin but should monitor for concentrated earners. Use actual redemption behaviour, not the face value of points, when you set the cap.

What would change our mind

We would reverse the recommendation if the programme's economics changed so that marginal contribution exceeded marginal liability across the eligible population. That could happen through a redemption value cut, an earn rate reduction, supplier-funded points, or a margin improvement in the capped categories. If a cap is not needed to preserve margin, keep earning uncapped and use burn incentives instead, because caps reduce member satisfaction without reducing a loss that does not exist. We would also change if regulation required caps on liability rather than on earning, in which case the mechanism would shift to expiry or redemption limits, not earning caps.

Related