Loyalty program economics: what a 5% earn rate really costs

Every loyalty app dashboard shows you the same friendly numbers: points issued, members enrolled, redemptions this month. What it does not show you is the P&L. A 5% earn rate is a 5% discount on every order that triggers a point, except the cost is deferred instead of taken today. It sits on your balance sheet as a liability until the customer redeems or the points expire. This post runs the full math for a brand doing $4M in repeat revenue: what the program actually costs, what breakage claws back, and how much incremental repeat revenue you have to prove before the program earns its place.

The math your loyalty app dashboard doesn't show you

Start with the headline number. A 5% earn rate on $4M of repeat revenue issues $200,000 of point liability every year. That is $4M multiplied by 5%. Each of those dollars is a promise to give a customer something in the future, and accounting treats it that way: it is a deferred cost, not a marketing metric you get to celebrate.

The reason this trips up so many operators is that the cost never appears as a line item they recognize. There is no invoice for "point liability." It shows up later, quietly, as margin erosion on redemption orders, or as a growing number in an accounting note nobody reads. When I talk to founders running a brand this size, the thing they keep saying is that the loyalty program "basically pays for itself," and when we pull the actual redemption data, it usually does not, or at least not in the way they assumed.

Think of the earn rate as a variable cost that scales directly with your repeat revenue. Grow repeat revenue from $4M to $6M and your gross point liability grows from $200,000 to $300,000 at the same 5% rate. That is fine if the program is driving the growth. It is a problem if the growth would have happened anyway and you are now paying 5% on all of it. The whole question of this post is which of those two worlds you actually live in, and whether you can prove it. That proof is exactly the exercise a fractional CFO runs during a program review.

What breakage is, why 25% is the retail benchmark, and why it isn't a windfall

Not every point gets redeemed. The share that never does is called breakage, and it works in your favor: a point that expires unused is a liability you never have to pay. Multiple 2024 to 2026 industry sources (Voucherify, RUSH, Rivo) converge on a blended retail breakage rate of about 25%, with healthy programs ranging from 5% to 30%.

That 25% is a blended average, and the blend hides a lot. Academic work (Goic 2025, in the International Journal of Research in Marketing) found that infrequent customers break at rates as high as 93%, while your most active, most valuable customers redeem almost everything. So the customers you least want to reward are the ones subsidizing the program through breakage, and the customers you most want to keep are the ones costing you full freight. Model the blended rate to size the liability, then watch it by segment to understand the risk. Loyalty program adoption rates vary by vertical, which means the segment mix you inherit will push your blended breakage above or below that 25% midpoint.

Apply 25% breakage to our $200,000 gross liability and expected redemption drops to $150,000. That is 3.75% of repeat revenue: a real, recurring margin line. Here is the accounting catch most small brands miss. Under ASC 606, breakage is not a lump-sum windfall you book the day points expire. It is recognized proportionally over the redemption period as other points get used. So even the good news arrives in slow motion.

Line	Amount	How it is calculated
Repeat revenue base	$4,000,000	Given
Gross point liability	$200,000	5% earn rate x $4M
Less breakage recovery	($50,000)	25% of gross liability
Net redemption liability	$150,000	3.75% of repeat revenue

The platform fee: what Smile, Yotpo, and LoyaltyLion actually cost a $4M brand

The point liability is the big number, but it is not the only one. You also pay a platform to run the program, and that fee scales with your order volume in steps, not smoothly. For a brand at $3M to $5M, the realistic range is $2,400 to $12,000 a year depending on platform and volume.

Smile.io Growth is $199 a month (about $2,388 a year) and covers up to 2,500 orders a month with VIP tiers included. Cross that order cap and you are pushed to Smile.io Plus at $999 a month, or roughly $11,988 a year. Yotpo's entry Loyalty Pro tier also starts near $199 a month but covers far fewer orders, so a brand at your scale likely lands on Yotpo Premium, estimated around $799 a month ($9,588 a year). LoyaltyLion Classic starts at $199 a month for 500 orders and scales more gradually, into the $549 to $729 range at higher volume.

The pattern to notice is the cliff. Smile.io stays flat and cheap right up until 2,500 orders, then quintuples. If your monthly order count is bumping against a tier cap, a good month can quietly move you into a fee bracket you did not budget. The pattern we see again and again is a founder who priced the program off the entry tier and got surprised by the renewal.

Platform	Recommended tier	Monthly	Annual	Order cap
Smile.io Growth	Best under 2,500 orders/mo	$199	$2,388	2,500/mo
Smile.io Plus	2,500-7,500 orders/mo	$999	$11,988	7,500/mo
Yotpo Loyalty Pro	Entry, likely light at $3M-$5M	$199	$2,388	~500-1,000/mo
Yotpo Loyalty Premium	Recommended at $3M-$5M	$799	$9,588	~3,000/mo
LoyaltyLion Classic	Entry	$199	$2,388	500/mo
LoyaltyLion Advanced	Best at scale	$549-$729	$6,600-$8,700	4,000/mo

The full P&L: total cost versus the breakeven lift

Now stack the two costs together. Net redemption liability of $150,000 plus a mid-range platform fee of $7,200 gives a total program cost of about $157,200 a year, before you credit back a single dollar of the extra revenue the program supposedly drives.

That $157,200 is the hurdle. To clear it, the program has to generate enough incremental repeat revenue that the margin on it covers the cost. At a 30% contribution margin, you need $157,200 divided by 0.30, which is $524,000 in additional repeat revenue. On a $4M base, that is a 13.1% lift. Not enrollment. Not points issued. A genuine, causal, 13% increase in repeat revenue that would not have happened without the program.

Scenario	Earn	Breakage	Net cost	Total cost	Revenue needed (30% CM)	Lift required
Conservative	3%	30%	$84,000	$91,200	$304,000	7.6%
Base case	5%	25%	$150,000	$157,200	$524,000	13.1%
Aggressive	7%	15%	$238,000	$250,000	$833,333	20.8%

Now weigh that against what the lift benchmarks actually say. Vendor numbers look spectacular: Yotpo's 2023 benchmark report shows redeemers with a 164% higher repeat-purchase rate and 88% more revenue per customer. But those compare redeemers to non-redeemers, a self-selected cohort, so they measure who joins, not what the program caused. A Harvard Business Review analysis of over 10,000 customers at a major retailer found most of the apparent lift was selection bias: members were already bigger spenders before they joined. The defensible, causal planning number from McKinsey and Rivo synthesis is 12% to 18%.

Line that up with the scenarios. The conservative program needs a 7.6% lift and clears the 12% to 18% band comfortably. The base case needs 13.1% and only works at the high end of the causal range. The aggressive program needs a 20.8% lift, above what almost any brand can prove. The higher your earn rate, the harder your program has to work, and the chart above shows the wall you hit.

A loyalty program is not a marketing perk with a monthly SaaS bill. It is a variable liability that scales with your repeat revenue, plus a stepped platform fee, minus whatever breakage you can defend. At a 5% earn rate you are underwriting a 13% repeat-revenue lift. If you cannot measure that lift, you are not running a loyalty program, you are running a discount you forgot to price.

When loyalty programs add margin: what the winners do differently

The brands that make the math work are not the ones with the most generous earn rate. They are the ones that treat redemption rate as a leading indicator and design around active members. A healthy program sees 20% to 30% of points redeemed per period by an engaged core, and that core is where the incremental revenue actually lives.

Three moves separate the winners. First, they measure incrementality properly, with a holdout group or matched cohorts, so they know the causal lift rather than the vanity lift. Second, they reward frequency, not just spend, because a program that only pays your biggest spenders is subsidizing behavior that already existed. Third, they avoid double-hit mechanics. One client running a loyalty program structured as discount codes against the earn rate flagged exactly this: "we don't want to take two hits," the earn-rate margin cost at issuance plus the discount cost at redemption. Pick one mechanic per order.

There is also a quieter lever most operators miss on the way out. Because breakage is recognized over time, winding a program down or restructuring it can release the remaining liability and create a short-term profit bump. As one operator put it in a client session, "what will happen in the short term is we will actually bump a bit of profit from removing these things this year." That is not a reason to kill a working program, but it does mean the accounting cost of an underperforming one is higher than the dashboard suggests, and the cost of exiting is lower.

The decision: launch, keep, or kill

Run your program through five questions before your next renewal. First, what is your current repeat-purchase rate baseline, and is it moving? Second, can you measure member versus non-member incrementally, or are you guessing? Third, does the program's total cost beat what the same budget would return in retention email, SMS, or a paid membership? Several operators we work with have found a simple paid tier does more with less: one running a $5-a-year subscription noted subscribers spend 1.5x the lifetime value of non-subscribers, a cheaper retention lever than a full points program. Fourth, are you carrying the correct liability on your books, or is your Shopify dashboard telling you a happier story than an accrual-basis P&L would? Fifth, if you cancelled the program tomorrow, what would you do with the budget, and what is that alternative worth?

The context for all of this is that loyalty liabilities are real enough that public companies disclose them in detail. Starbucks carries about $1.75B in combined stored-value and loyalty liability and recognized $222.4M of breakage revenue in FY2025. Delta's SkyMiles sits at $8.8B in deferred revenue; Marriott's Bonvoy at roughly $8B, where a single percentage-point shift in the breakage estimate moves the liability by about $50M. Your program is smaller, but the mechanics are identical, and so is the discipline required. The winners run the numbers before they run the promotion.

Sources and methodology

The core P&L model is built from public benchmarks, not proprietary panel data. The $4M repeat revenue base and 5% earn rate are the modeling premise. Gross point liability, net redemption liability, platform fees, and breakeven lift are all derived from that premise combined with the published breakage, pricing, and lift benchmarks cited below. Figures are directional planning numbers, not a specific brand's actuals.

Breakage benchmarks come from industry and academic sources. The 20% to 30% retail range (25% midpoint) is from the Voucherify loyalty breakage glossary and corroborating 2024-2026 industry write-ups. The finding that infrequent customers break at up to 93% while active customers approach zero is from Goic (2025) in the International Journal of Research in Marketing.

Platform pricing is from public pricing pages and third-party comparisons. Smile.io tiers (Growth $199/mo, Plus $999/mo) are from the Smile.io pricing page and LoyaltyLion Classic from the LoyaltyLion pricing page. Yotpo Premium ($799/mo) and LoyaltyLion Advanced pricing are estimated from third-party comparisons because Yotpo's full pricing is not publicly rendered; confirm with a direct quote before budgeting.

Repeat-purchase lift benchmarks separate vendor claims from causal evidence. The 164% and 88% redeemer figures are from the Yotpo Loyalty Program Benchmarks Report (Aug 2023) and reflect a self-selected redeemer cohort. The defensible 12% to 18% causal range is from McKinsey and Rivo synthesis; the selection-bias caveat is from a Harvard Business Review analysis of a large-retailer customer base. Antavo's 2024-2026 statistics put average program ROI at 4.8x to 5.3x with 74% of members disengaging within two months.

Public-company loyalty liabilities are drawn from SEC filings. Starbucks liability and breakage figures are from its FY2025 Form 10-K; Delta SkyMiles deferred revenue from its 2024 10-K; and Marriott Bonvoy deferred revenue and breakage sensitivity from its FY2025 Form 10-K.

The accounting treatment follows ASC 606 guidance. Loyalty points that convey a material right are a separate performance obligation, deferred at standalone selling price and recognized on redemption or expiry, with breakage recognized proportionally over the redemption period.

Frequently asked questions

how much does a 5% earn rate actually cost me per year in real dollars?

On $4M of repeat revenue, a 5% earn rate issues $200,000 in point liability a year. After the 25% retail breakage benchmark, expected redemption is $150,000. Add a platform fee near $7,200 and the program costs about $157,000 a year before any repeat-purchase lift is credited back.

what's a typical breakage rate for a shopify ecommerce brand?

The retail benchmark is 20% to 30%, with 25% as the consensus midpoint. But it varies wildly by cohort: infrequent customers can break at over 90%, while your most active buyers redeem almost everything. Model the blended rate, then watch it by segment.

how are loyalty points treated on my balance sheet, are they a liability?

Yes. Under ASC 606 a point that gives a customer a material right to future goods is a separate performance obligation. You defer a portion of each sale as a liability and recognize it as revenue only when points are redeemed or expire. Most cash-basis Shopify brands are not accruing this at all.

when does a loyalty program add margin vs destroy it?

It adds margin when the incremental repeat revenue it drives exceeds the net point liability plus the platform fee. On the base case that means clearing roughly $524,000 in extra repeat revenue. If you cannot measure member versus non-member incrementally, you cannot know which side of the line you are on.

is a 5% earn rate too generous, what earn rate should i use?

For most brands at 30% to 40% gross margin, 3% to 5% is the workable range. At 7% earn and low breakage the program needs a 20%-plus repeat lift to break even, which almost no brand hits. If your margins are thin, start at 3% and raise it only if you can prove the lift.

how do i calculate the breakeven repeat purchase lift my loyalty program needs?

Take your total program cost (net point liability plus annual platform fee) and divide by your contribution margin. That is the incremental repeat revenue you need. Divide that by your repeat revenue base to get the lift percentage. On the base case it is $157,200 divided by 0.30, or $524,000, about a 13% lift.

loyalty program vs discount codes: which is cheaper for driving repeat purchases?

Points defer the cost and let breakage claw some of it back, which is cheaper on paper than a same-day discount. The trap is stacking both: earning points and then letting customers redeem them as a discount code is a double hit. Pick one mechanic per order, not two.

what happens to my p&l when customers don't redeem their points?

Unredeemed points are breakage, which reduces your net liability. Under ASC 606 you recognize that breakage as revenue proportionally as other points get redeemed, not as a one-time windfall. If you wind a program down, you may recognize a short-term profit bump as the remaining liability releases.