Loyalty glossary · 2. Accounting and finance (25)
Liability Per Member
Liability per member is the average deferred obligation a loyalty programme carries for each individual member, calculated by dividing the total outstanding points liability by the number of members.
Liability per member is the first number an operator reaches for when asked how much its promise costs. It divides the total deferred liability on the balance sheet by the member count, giving a tidy dollar figure per head. The trouble is that it is an average, and averages in loyalty programmes hide more than they reveal.
The trap is treating a low liability per member as a sign of health. A programme can lower that figure by acquiring thousands of inactive members who never earn or redeem, or by writing off old balances that should have been chased. Neither action creates value, but both flatter the metric. A high figure can also be misleading if a few accounts hold the bulk of the liability.
Run the numbers on a partner programme to see the distortion. A scheme with a 20 million dollar liability and 500,000 partners reports a 40 dollar liability per partner. If 80 percent of those partners are inactive and each holds only 1 dollar, the remaining 100,000 partners carry a 196 dollar liability per partner. The headline average is ten times smaller than the burden on the members who actually engage.
The metric also fails on redemption risk. A liability of 20 dollars per member means nothing without knowing how many of those dollars are attached to points that will expire in 30 days versus points that sit in dormant accounts for 5 years. Ageing the liability by cohort is the only way to get an honest number, and almost no programme reports it.
ASC 606 and the equivalent accrual rules require a best estimate of the liability, not a per member average. Any operator that presents liability per member as a disclosure is dodging the real question: what is the expected redemption of that liability, and when? The per member figure is a lazy denominator, and it should not drive accrual decisions or breakage assumptions.
Use liability per member as a first glance, not a decision input. The moment it is compared across programmes or tracked over time without a distributional view, it becomes a tool for hiding risk. A programme that reports a stable 35 dollar liability per member while its active member liability doubles is not stable; it is just averaging away its problem.