Breakage

Breakage is the portion of a stored-value obligation — loyalty points, store credit, or gift-card value — that a company reasonably expects will never be redeemed. Per Milliman it is calculated as the activation amount less the ultimate redemption value; per Voucherify it is the gap between points earned and points redeemed, and it sits on the balance sheet as a financial liability as well as serving as an engagement diagnostic. It is the unredeemed-credit failure mode named repeatedly across the vault's loyalty cluster (Loyalty Programs, Tiered Loyalty Programs, Paid Loyalty / Premium Membership). The central tension the sources surface is whether breakage is a CFO-side "win" (lower program cost) or a CX-side red flag (disengagement that predicts churn). All quantitative figures this run are vendor- or aggregator-sourced; no independent/academic benchmark was found.

Firewall: every claim is what a source reports. See ../../CONTEXT.md Rule 1.

How it works

A stored-value sale or points-earn event creates a contract liability (deferred revenue) — the company owes future goods/services. Three states govern the lifecycle (Milliman): activation (liability created), redemption (amount used), and breakage (the residue never redeemed, which is eventually released to revenue subject to the rules below).

Key terms

Term	Meaning (per sources)
Activation	Liability created when a card is bought / points are earned (Milliman)
Redemption	Value actually used by the customer (Milliman)
Breakage	Value reasonably expected never to be redeemed (Milliman; Voucherify)
Contract liability / deferred revenue	How the unredeemed obligation is first booked under ASC 606 (RevenueHub)
Escheatment	State unclaimed-property claim on the unredeemed balance, ahead of breakage revenue (RevenueHub; Alston & Bird)
Proportionate method	Recognising breakage revenue in proportion to actual redemptions, not upfront (RevenueHub)

Breakage rates (as-of 2026-06-27 — all vendor/aggregator-sourced, volatile)

Loyalty, retail: ~25% (Voucherify); CPG retail 20–30% (Brandmovers).
Loyalty, travel & airlines: 30%+ (Voucherify); B2B and travel up to 70–85% (Brandmovers).
"Healthy" band: 5–30%; >40–50% signals rewards aren't compelling or redemption is too hard (Voucherify).
Gift cards (illustrative worked examples, not measured): ~10% (Milliman redemption-curve example) and 20% (RevenueHub ASC 606 example).
Bond 2026 (Canada): 28% of members redeem once a year or less; ~$13–15B in unredeemed loyalty points outstanding.

Retail loyalty breakage benchmark — Voucherify ~25% [voucherify.io] vs Brandmovers 20–30% [brandmovers.com]. Roughly consistent, but both vendor estimates without primary survey data.

Gift-card worked-example rates — Milliman ~10% [milliman.com] vs RevenueHub 20% [revenuehub.org]. Both are illustrative assumptions, not measured market rates.

[!unverified] Aggregate US gift-card figures (~5–15% of sales unredeemed, ~6% of cards never used, ~$21B unredeemed) come only from secondary aggregators (savemycent/Hubifi/Enjovia) with no traceable primary source.

The CPG 20–30% / travel 70–85% figures originate in Brandmovers' 2022 text; the Milliman example is 2022. Included as the best-attributed rate sources found; no fresher equivalent surfaced. Re-check before relying on the exact numbers.

Accounting treatment (ASC 606 / IFRS 15)

Per RevenueHub (describing the unchanged ASC 606 standard):

Stored-value proceeds are booked first as a contract liability (deferred revenue).
If breakage can be reasonably estimated and the entity expects to be entitled to it, it reduces the standalone selling price for estimated breakage and recognises breakage revenue in proportion to actual redemptions (the proportionate / redemption-pattern method). Recognising breakage upfront is inappropriate.
If the entity does not expect to be entitled to breakage, it recognises the revenue only when the likelihood of future redemption becomes remote.
Escheat exception (ASC 606-10-55-49): where the unredeemed amount is owed to the state, the entity books a liability payable to the government instead of breakage revenue.
A portfolio approach is permitted — aggregate estimation across many contracts. Cited examples: The Cheesecake Factory recognises breakage over ~3 years in proportion to historical redemption; Target recognises it over time in proportion to actual redemptions (cards don't expire).

Per Milliman, estimation in practice uses actuarial reserving methods (multiplicative/ loss-development, additive, cede-ratio) by activation period; auditors often require third-party support; seasonality (Q4 gift-card sales redeemed in Q1) and channel/denomination mix materially shift the redemption curve.

[!unverified] IFRS 15 treatment confirmed only as "aligned with ASC 606" via a secondary snippet; no IFRS primary fetched. The mechanics above are US GAAP.

Escheatment / unclaimed-property law

Most state unclaimed-property laws have gift-card provisions, but 37 states expressly exempt gift cards (incl. CA, IL, FL, OH, PA, TX), often conditioned on no expiration and no inactivity fees; 14 jurisdictions (incl. DC) do require escheatment (Alston & Bird, as-of 2025-10-01).
2025 saw no escheatment-specific gift-card legislation; most activity was fraud-prevention (NE, NJ, MD, DE, NY, RI); pending California SB 22 would raise the cash-back redemption threshold to $15 (Alston & Bird).
Federal Reg E / Credit CARD Act (12 CFR §1005.20) sets a 5-year minimum before underlying gift-card funds can expire and limits dormancy fees to one per month after 12 months of inactivity (CFPB).

[!unverified] Washington State reportedly introduced bills expanding the gift-certificate definition to include loyalty points (2024–2025), which would bring points into escheatment — snippet only, not verified at primary. If it stands, it would directly raise the cost of loyalty breakage in that state.

Program economics — is breakage good or bad?

The sources frame a structural tension (Voucherify):

CFO view: high breakage is a "win" — lower program cost and a smaller balance-sheet liability. Some brands engineer it via high redemption thresholds or short expiry windows.
Growth/CX view: high breakage signals disengagement, predicts churn, and erodes the switching costs loyalty is meant to build. Voucherify cites "redeemers" churning 8–20% less than non-redeemers (unsourced).
Brandmovers argues the negative impact of high breakage outweighs the saved fulfilment cost, and its 2026-trends section claims leading brands now treat breakage as a CX KPI not a cost lever — moving to earn-and-burn models, micro-redemptions, AI-personalised redemption prompts, and proactive expiry communications; "programs with intentionally high breakage rates are increasingly viewed as outdated."
This good-vs-bad framing connects to Tiered Loyalty Programs's design rule: route non-transactional actions to status credits (non-redeemable, no liability) rather than reward currency, precisely to limit breakage liability.

"Designing for breakage" — vendor sources uniformly frame engineered breakage as bad/outdated [voucherify, brandmovers] while conceding CFOs treat it as a win. Near-unanimous vendor narrative with a clear conflict of interest (they sell redemption/engagement tooling); no independent CFO-side defence of breakage was found this run.

What practitioners report

Nothing this run — both practitioner streams were down. The reddit-research MCP and the Apify transcript actor were not connected, so there is no operator-side observed-breakage data or shopper counter-narrative on expiring points. Carried as a gap.

Gaps

No fashion/apparel-specific breakage figure (closest: retail/CPG and restaurants). Carries the running UNIQLO-Europe / fashion data gap.
No independent/academic benchmark — all rates are vendor or aggregator.
IFRS 15 specifics and airline "ghost economy" figures unverified / not fetched.
No practitioner counter-narrative (Reddit + YouTube down).
Frontier links left dangling: Gift Cards, Deferred Revenue, Escheatment, Zero-Party Data, Loyalty Programme UX.