Decisions · Strategy and governance

Should you sunset a loyalty programme, and how should you end it

Yes, sunset a loyalty programme when it no longer changes member behaviour enough to pay for its own liability and operations. End it by fixing an accrual stop date, honouring existing points for a limited redemption window, and writing off the remaining liability in one accounting period.

The decision

The decision is whether to close a loyalty programme and, if so, how to end it without creating a customer revolt or an accounting surprise. The mechanism is the outstanding points balance: every point issued is a promise to deliver future value, so it sits on the balance sheet as a liability. Sunsetting is therefore a liability extinguishment exercise wrapped in member communication, not just a marketing retreat. A programme that is not changing member behaviour should be closed, but the close must give members a clear end date and a fair redemption window.

What it costs you

Closing a loyalty programme costs you the remaining liability, the margin on points that are redeemed, the engineering time to switch off earning and redemption channels, and member trust if the wind-down feels abrupt. Consider an illustrative programme with 150 million points outstanding at one cent per point. The liability is 1,500,000 dollars. If you stop accrual and allow a 120-day redemption window, and historical behaviour suggests 35 percent of points will be redeemed, you will issue 52.5 million points of reward value, which costs 525,000 dollars in margin and goods. You write off the remaining 97.5 million points, or 975,000 dollars, as a liability reduction. Separately, two engineers working six weeks at 40 hours per week and 150 dollars per hour is 72,000 dollars. Customer service overtime for 200 extra hours at 25 dollars per hour is 5,000 dollars. The total cash and margin outlay is 602,000 dollars against a liability that was 1,500,000 dollars. The accounting gain is real, but it does not capture the loss of member goodwill.

How to decide

Apply this test to your own programme. Start with the annual increase in the outstanding points liability. Add the fully loaded cost of running the programme: platform fees, fraud, service time, and communication. Compare that total to the incremental gross margin you can attribute to members who earn and redeem points. If the programme's attributable margin is lower than its liability growth plus operating cost for two consecutive years, sunset it. Be strict about attribution: do not credit the programme with purchases a member would have made anyway. If you cannot show that the programme changes behaviour, you are running a discount scheme with extra steps. When you decide to close, set the accrual stop date before you announce it, then give members at least 90 days to redeem, and extend that to 180 days for any member with more than 20,000 points. That gives high-value members time to use the value without allowing the liability to linger.

What would change our mind

We would reverse this recommendation if your programme is still acquiring members at a rate that lowers average liability per member and if a controlled test using existing member data shows point-earning members have a repeat purchase rate at least 15 percentage points higher than matched non-members over 12 months. In that case, the programme is doing real behavioural work, and the right move is to keep the programme but renegotiate the redemption terms or revalue the point rather than close it. We would also change our mind if a regulator has signalled that unilaterally cancelling points would be treated as an unfair contract term; then the only acceptable route is to run the programme down through natural redemption and communicate an end to accrual well in advance.

Related