Decisions · Economics and liability

What should you do when the programme is cannibalising margin

Cut the earn rate to zero on any category where loyalty points push net margin below your floor. Keep the programme alive on purchases that clear that floor, and convert the saved liability into a visible member discount rather than more points.

The decision

The decision here is not whether to shut the whole programme; it is whether to reset the earn rate on the specific baskets that are destroying margin. A loyalty liability is a variable cost attached to the basket at the moment of accrual. If you wait until redemption to manage that cost, you are trying to recover a price decision after the sale. The mechanism is the accrual rate per category. When a category earns points at a flat rate while its gross margin is thin, the points convert an acceptable sale into an unacceptable one. So the decision is to treat loyalty cost as a pricing line, not a marketing overhead.

What it costs you

The recommended path costs engineering time to segment earn rules, and it costs some member trust because removing points from a familiar category reads as a devaluation. But the larger cost of doing nothing is permanent margin erosion. Suppose a member buys a basket priced at 100 dollars at 30 percent gross margin, which is 30 dollars. At 5 points per dollar, the basket earns 500 points. If each point costs 1.2 cents to fund, the liability is 6 dollars. Margin after loyalty cost falls to 24 dollars, or 24 percent. Illustrative as it is, the arithmetic shows the point programme can consume a fifth of the gross margin before any redemption friction.

Now cut the earn rate to zero on that category. On 1,000 such baskets a month, the saved liability is 6,000 dollars a month and 72,000 dollars a year. If you reinvest 40 percent of the saved liability into a member discount of 2 dollars on the same baskets, that discount costs 2,000 dollars a month, leaving 4,000 dollars a month in recovered margin. The member sees a tangible price cut of 2 dollars instead of a future points promise, and the programme stops buying unprofitable volume.

How to decide

Pull your last three months of basket-level margin data. For each product category, compute margin after loyalty cost as gross margin dollars minus the product of basket value, earn rate, and cost per point. Flag any category where this figure falls below your floor margin in two consecutive monthly cycles. For those categories, set the earn rate to zero. For categories that clear the floor by at least five percentage points, keep a reduced earn rate. Then apply a single test to the borderline cases: ask whether a member would still buy that basket without points. If the answer is yes, the points are not creating the purchase; they are only cost. Remove them.

What would change our mind

We reverse this view when the cannibalising categories are retention drivers, not transaction drivers. If your own cohort data show that members who earn on those categories have a repeat purchase rate or lifetime value high enough to offset the margin loss within the period you model, then keep the earn rate and recover margin elsewhere, such as through a higher redemption threshold or by shifting funding to supplier partners. That condition is specific: the programme must be able to point to a measured revenue effect, not a brand halo, because a loyalty programme that cannot show that effect is just a discount with extra steps.

Related