Every loyalty program is, at its core, a currency system. Points are issued at a rate (the earn side) and redeemed at a rate (the burn side), and the spread between the cost of issuing points and the cost of honoring redemptions is what determines whether the program is a net contributor to the business or a net cost. Getting the earn/burn ratio right is the most consequential economic decision in loyalty program design.
This is also the decision that's most frequently made on intuition or by copying what a competitor appears to be doing, without actually modeling the economics. This post offers a framework for doing the modeling.
The Core Earn/Burn Equation
The fundamental relationship:
Earn rate: Points issued per dollar of qualifying spend. Example: 1 point per $1 spent.
Burn rate: Dollar value per point redeemed. Example: 100 points = $1 discount.
Effective discount rate: (earn rate × burn rate) = the percentage of spend that comes back to the member as reward value. At 1 point/$1 and 100 points/$1, the effective discount rate is 1%.
Expressed differently: if a member spends $100, earns 100 points, and eventually redeems those 100 points for $1 off a future purchase, your loyalty program is providing a 1% effective discount on qualifying spend. The question is: does that 1% drive enough incremental behavior — more visits, larger baskets, reduced churn — to justify the cost?
For retail programs, effective discount rates generally cluster in the 1%–3% range for standard earn. Below 1% tends to produce low engagement because the perceived value is too thin for members to think about. Above 3%–5% on standard earn starts to erode margin unless the program is driving significant lift in basket size or visit frequency. Coalition program structures sometimes run higher effective rates because the cost is shared across multiple merchants.
Category-Differentiated Earn Rates
A flat earn rate across all categories is the simplest design, but it doesn't reflect the margin reality of a retail business. A sporting goods chain earns a very different gross margin on a pair of running shoes (typically 40–50%) versus a protein supplement (30–40%) versus a clearance-priced bike helmet (under 20%). Providing the same loyalty earn rate on all three products means the program's effective cost is much higher on low-margin items than on high-margin ones.
Category-differentiated earn rates address this: high-margin categories earn at a higher rate (or at base rate), low-margin categories earn at a reduced rate or are excluded from program earnings. The tradeoff is communication complexity — if members notice that the category they most often buy earns fewer points, the perceived generosity of the program falls, even if the aggregate economics improve.
The implementation path that minimizes member confusion: establish a clear standard earn rate as the "base," and use category overrides only to increase the earn rate (bonuses on selected categories) rather than decrease it. Members who never buy the low-earn categories never notice. Members who exclusively buy low-margin categories may churn — but those are also the members with the most adverse economics for the program, so their churn may be economically neutral or even positive.
We're not saying excluding low-margin categories from earning is universally the right call. We're saying the decision should be made with explicit margin modeling, not by copying a competitor's program structure without knowing their margin profile.
Modeling the Cost of a Point
The loaded cost of a loyalty point is not simply the face value of the redemption it enables. To model the full cost, you need:
- Expected redemption rate: What fraction of issued points will ultimately be redeemed? This is your breakage-adjusted redemption rate. If you issue 1,000 points and expect 30% breakage, 700 points will be redeemed. The liability you need to reserve is for those 700 points, not the full 1,000.
- Redemption catalog cost: What is the cost-to-program of the average redeemed reward? For dollar-off discounts, this is the face value. For product rewards, this is the wholesale cost of the item. If your catalog is primarily dollar-off, the cost per point is straightforward. If it's mixed, the weighted average cost per point redeemed depends on what members are actually redeeming for — which you can only accurately model after the program has been running for a full redemption cycle.
- Platform and operational cost: The licensing cost of your loyalty platform, customer service overhead for loyalty-related contacts, and engineering maintenance should be amortized across your qualifying spend volume to produce a per-dollar-of-spend overhead rate. This is often excluded from earn/burn modeling and then surprises operators when the total program cost is higher than the redemption liability alone.
A fully loaded model: if your earn rate is 1 point/$1, your redemption rate (adjusted for breakage) is 70%, your catalog redemption costs an average of $0.009 per redeemed point, and your platform/operational overhead is $0.002 per point issued — the total loyalty cost per dollar of qualifying spend is approximately (1 × 0.70 × 0.009) + (1 × 0.002) = $0.0083, or about 0.83% of qualifying spend. Your 1% effective discount rate is actually costing you 0.83% in total loaded cost — the 0.17% gap is the net positive economic position from breakage.
Promotional Multipliers and Liability Spikes
Bonus earn events — 2x, 3x, or 5x points on qualifying purchases during a promotional period — are an effective member engagement tool, but they create a deferred liability spike that your redemption catalog needs to be ready to absorb. Members who earn a large bonus point balance will redeem it over the following weeks or months.
The modeling trap: a 3x weekend event on a $500,000 gross sales day at an 87-location outdoor retail chain would issue 3x the normal points on that day's qualifying spend. If standard earn is 1 point/$1 and average qualifying spend per transaction is $65, a 3x day issues 3 points per dollar. Across 87 locations running 300 transactions each, that's 87 × 300 × 65 × 3 = 5.1 million points in a single day — three times the normal daily issuance. Those points will start hitting your redemption catalog 4–8 weeks later. If your breakage rate is 30%, you're looking at 3.6 million points redemption-bound in the post-event period. At $0.009 per redeemed point in catalog cost, that's approximately $32,400 in deferred redemption cost from a single promotional weekend.
Whether that's economically justified depends entirely on what the 3x event drove in incremental basket size, store traffic, and new member acquisition. Without measuring the lift against the baseline, you can't tell if the promotional economics were positive. This is why promo-period analytics is one of the highest-value reporting capabilities in a loyalty platform — it closes the loop on whether you actually got the revenue lift you paid for.
Adjusting Earn/Burn Over Time
Earn rates and redemption values are not set-it-and-forget-it parameters. They should be reviewed annually at minimum, and adjusted when the program's engagement metrics suggest the current ratio is either too generous (causing economic strain) or too stingy (causing disengagement).
Signs the earn rate may be too high: redemption rate above program targets, per-member program cost above budget, high-frequency low-margin purchasers dominating your active member cohort. Signs the earn rate may be too low: high breakage on members who have been in the program for 6+ months, low first-redemption rate, declining active member percentage year-over-year.
Adjusting earn rates mid-program is politically complex — members notice reductions. The cleanest path is adjusting through catalog pricing (increasing the point cost of catalog items, which effectively reduces the burn rate without touching the earn rate) or through new category multipliers that increase earning in strategic categories while the overall base rate stays flat. This preserves the "you earn the same on every purchase" communication while the effective economics shift.
Revlance's program economics configuration is designed to support these adjustments through parameter changes rather than schema changes — earn rates, multipliers, and catalog pricing are all API-configurable without a code deployment, so operators can test adjustments and measure their effect before committing to a program-wide change.