Why loyalty programmes work at all

The goal gradient effect, first documented in rats running toward food and later replicated in coffee card loyalty programmes, establishes the foundational mechanism. People accelerate toward a goal as they perceive themselves getting closer to it. Kivetz, Urminsky and Zheng’s 2006 study found that coffee loyalty members purchased more frequently as they approached a free drink — and that this acceleration predicted retention and re-engagement after the reward was redeemed.

The endowed progress effect extends this. Members given a head start — a card already partially stamped before they’ve made a purchase — engage more than members who start from zero, even when the total effort required is identical. The psychological experience of already being underway activates goal-gradient acceleration earlier. Framing rewards as progress rather than accumulation increased programme participation by 20% in documented trials — not because the reward changed, but because the perceived proximity to it did.

The failure condition is the mirror. A programme where the reward feels permanently out of reach produces none of this. If members cannot perceive movement toward the goal, the acceleration mechanism never activates, and the programme generates no behavioural change at all.

How loss aversion makes status sticky

Loss aversion — the well-established finding that losses are weighted roughly twice as heavily as equivalent gains — is the mechanism behind tier stickiness. Once a customer achieves a status level, they develop loss aversion around maintaining it. The fear of losing Gold status motivates behaviour that the prospect of gaining Gold status alone would not.

Airlines have engineered this deliberately for decades. The threat of status tier expiry does not merely remind customers of a benefit they could lose — it activates the asymmetric emotional weighting of loss that makes defensive behaviour disproportionately motivated. Research finds that 37% of customers report increasing spending specifically to maintain tier status. The gradient of avoidance is steeper than the gradient of approach.

The failure condition: loss aversion requires a meaningful object to attach to. When tier benefits are diluted as membership grows, or when the tier ceases to feel genuinely exclusive, there is nothing worth protecting. The mechanism that drove stickiness dissolves with the perceived value of what was at stake.

The structural reason programmes eventually stop working

Hedonic adaptation is the mechanism that undermines every mature loyalty programme. The emotional impact of rewards fades with repetition. Today’s surprise benefit becomes tomorrow’s baseline expectation. The benefit that produced genuine delight at launch is processed as entitlement by a long-tenured member — and any reduction, however modest, is experienced as a loss relative to the new baseline rather than as a return to the original state.

Research specifically examining loyalty programmes found that customer tenure is a key moderating variable: longer-tenured members are more susceptible to hedonic adaptation than newer ones. The programmes that have retained customers longest are therefore the ones most exposed to this failure mode. Starbucks experienced this directly in 2016 when restructuring its Rewards programme. Long-tenured members experienced the change as a loss relative to their established psychological baseline, and complaint volume was disproportionate to the actual financial impact on most members — because the reference point, not the absolute level, was what mattered.

When loyalty rewards destroy what they reward

The overjustification effect introduces a fourth and underappreciated failure condition. When external rewards are applied to behaviours a customer was already intrinsically motivated to perform — recommending the brand, writing reviews, participating in community — the internal reason for the behaviour shifts from “I do this because I want to” to “I do this for the reward.” When the reward is reduced or removed, the intrinsic motivation has already been displaced and does not automatically return.

Loyalty programmes that reward organic word-of-mouth or community participation risk converting freely given behaviour into behaviour that is now dependent on payment. The programme has trained customers to need an incentive for something they once did voluntarily. The boundary condition the research supports is specific: extrinsic rewards work best applied to behaviours customers would not perform intrinsically — completing onboarding steps, providing structured feedback. They are most damaging when applied to behaviours that were already self-sustaining.

Why most programmes generate transactions but not loyalty

The attitudinal versus behavioural loyalty distinction explains the gap between programme membership numbers and genuine retention. Behavioural loyalty is repeat purchase driven by incentive structure. Attitudinal loyalty is emotional commitment that persists when a competitor offers a better deal. Most loyalty programmes are designed to generate the first while assuming it will produce the second. It doesn’t.

Eighty-one percent of consumers join loyalty programmes. Forty-nine percent actively use them. Point accumulation mechanics do not generate attitudinal commitment. Only genuine perceived value, identity alignment, and relational investment do. The programmes that produce both types of loyalty — Amazon Prime being the canonical case — do so through structures that generate investment, commitment-consistency, and identity-level attachment, not just transactional reward.

A book worth reading alongside this

The Loyalty Leap by Bryan Pearson is the most directly applicable treatment of loyalty programme design from a practitioner who spent decades running one of the world’s largest data-driven loyalty operations. Pearson’s account of what actually drives genuine customer commitment — as distinct from transactional repeat purchase — addresses the attitudinal versus behavioural loyalty gap the article identifies directly. His treatment of why programmes that generate impressive membership numbers consistently fail to generate the emotional attachment that prevents switching provides both the diagnostic framework and the practical design principles for building something that works beyond the first redemption cycle.

Have questions about this article?

If any part of this article raised questions you want to explore further, courbot.co is built for exactly that. It is courben.co’s AI assistant, designed around the psychology of entrepreneurship. Ask it anything from this article.

This article is for educational and informational purposes only. Sources: Kivetz, R., Urminsky, O. & Zheng, Y. (2006), Journal of Marketing Research, 43(1). Nunes, J.C. & Drèze, X. (2006), Journal of Consumer Research. Kahneman, D. & Tversky, A. (1979), Econometrica. Fu, X. & Ren, X. (2023), Nankai Business Review International.