Decisions · Risk and fraud

Should you investigate incentive leakage

Investigate incentive leakage when redemption cost per point varies by channel or partner, or member service complaints point to missing value. A one-time audit of issuance, redemption and expiry paths identifies the gap before it becomes a permanent margin and trust problem.

The decision

You are deciding whether to spend scarce audit time tracing how value leaves the programme, rather than accepting observed redemption and issuance as the cost of doing business. The mechanism underneath is not a single fraud event. It is the gap between what the programme's rules intend to cost and what members, partners and internal systems actually draw. Investigating means mapping each batch of points from issuance through redemption, transfer or expiry and comparing the actual cost per point against the liability model. A programme with variable redemption costs or partner-settled redemptions should run that mapping at least once.

What it costs you

Investigating leakage costs engineering and analytics time, and it can force recognition of liability you had not booked. The real cost is not the audit report; it is the two to six weeks of data extraction, reconciliation and partner queries. An illustrative calculation shows the trade-off. Say a programme issues 10 million points per month and books a blended redemption cost of 0.8 cents per point. Monthly cost is 80,000 dollars. If leakage equals 0.5 percent of issued points, that is 50,000 points per month. At the blended rate, leakage costs 400 dollars per month, or 4,800 dollars over 12 months. If the same 50,000 points redeem through a partner channel at 3 cents per point, they cost 1,500 dollars per month, or 18,000 dollars over 12 months. An investigation costing one analyst six weeks at 15,000 dollars and half an engineer for two weeks at 4,000 dollars totals 19,000 dollars. That is recovered after 13 months of high-cost leakage, because 13 times 1,500 is 19,500 dollars. The larger cost may be finding that a partner has been settling at a rate your ledger did not reflect.

How to decide

Apply the reconstruction test. For the last three complete months, ask whether you can account for every 100,000 points issued: how many were redeemed, expired, transferred or remain outstanding, and what the actual cost per point was in each channel. If that reconstruction takes longer than five business days or cannot be produced from existing data, investigate incentive leakage now. If the reconstruction is immediate and the gap between the cheapest and most expensive redemption channel is under 0.5 cents per point, an investigation can wait until a new partner or transfer feature is added.

What would change our mind

We would reverse this recommendation for a closed-loop programme with a fixed price catalogue, no partner-settled redemptions, and no point transfers or pooling before expiry. In that structure leakage is capped by the catalogue price, the liability per point is known, and the only failure mode is an accounting error rather than a mismatched incentive. An audit then costs more than the maximum avoidable leakage it can find, so the correct decision is to rely on ordinary reconciliation instead of a dedicated investigation.

Related