Redemption Momentum

What it is

A common reading of loyalty data is that customers buy more as a reward gets closer, because the threshold pulls them toward it. Dorotic, Verhoef, Fok and Bijmolt asked a different question. If members choose how much to redeem and when, does the decision to redeem change what they buy, even without a threshold to reach or points about to expire?

They used weekly data on three thousand and ninety-four members of a nationwide coalition programme in the Netherlands, followed for one hundred and eighty-four weeks. Members earned points at more than ten partners across sectors that included grocery, fuel, insurance and travel. The sample was made up of members with a minimum number of purchases and at least one redemption.

Their finding was that redeeming, even a small fraction of the balance, raised purchasing in the weeks before and after. They call it redemption momentum. About seventy per cent of redemption decisions were made before the purchase that week, which fits the idea that the decision itself drives the earlier rise. Members redeemed about a quarter of their balance on average, so the pull of a nearing reward explains only a small share of it, in the authors' reading.

The effect faded over the following weeks. It varied with how long someone had been a member, with age and income, and with the number of direct mailings received.

In effect

In research terms, the paper separates two things that look alike in the data: buying more because a reward is close, and buying more because a decision to redeem has been made. That distinction is what the name is for.

In commerce, the finding bears on a worry that redemption costs a firm a customer. In this programme, purchasing rose around redemption instead of falling, though the programme belonged to a coalition of partners and not to a single retailer.

In everyday life, it fits the person who finally uses their balance and finds they shop more around that week. That is a plain illustration of the pattern, not a documented case.

What it does not say

It cannot say what happens to a member who never redeems, which is the group of most interest to anyone asking why points sit unspent. The sample held only members with at least one redemption.

It is observational. Members who redeem may differ from those who do not in ways the data do not capture, and the data the paper had were limited to such things as tenure, age, income and mailings.

It rests on one coalition programme in one country. For this page the published abstract and a summary of the paper were read, and the full methods were not.

How it differs from the goal gradient

The goal-gradient account, tested on customers by Ran Kivetz and colleagues in 2006, says people speed up as a reward gets closer. In their café programme, customers bought coffee more often as the free one came near. That is a claim about distance to a threshold, and it predicts a rise in buying before a reward is earned and a reset afterwards.

Redemption momentum is offered as a different reading. The programme Dorotic and colleagues studied let members choose how much to redeem and when, so there was no single threshold to approach. About 70 per cent of redemption decisions were made before that week's purchase, and members redeemed about 26 per cent of their balance on average. The authors read that as leaving only a small share of the rise before redemption to the pull of a nearing reward, and suggest that the decision to redeem is itself associated with buying more.

Who showed it more

The effect was not uniform. It was weaker among longer-standing and higher-income members and stronger among older ones, and it faded over the weeks that followed a redemption. A member who redeemed and then stopped buying would not appear in a pattern that fades, which is one reason the observation is hard to turn into a rule.

What a firm could conclude

A firm that fears redemption costs it a customer has a result pointing the other way in this programme. The caveats are the ones already stated: every member in the sample had redeemed at least once, so the study cannot say what happens to someone who never does, and the coalition setting differs from a single retailer. The finding is a reason to measure, not a reason to assume that encouraging redemption will raise sales.

---