Costing Tier Benefits: What Free Shipping, Early Access and Free Returns Really Cost Per Member

The short version: Price every tier perk at marginal cost per member per year, then keep only the perks the tier's own incremental margin funds. If the math funds zero perks, the tier list is wrong — not the budget.

Key takeaways

• Price perks at marginal cost, never retail value — retail value is what the perk feels worth, marginal cost is what it costs you.
• Per member per year is the costing unit; a perk's cost scales with how often qualified members actually order, not with averages.
• The budget is the tier's incremental gross margin — what members in that tier generate above what they'd spend anyway.
• Early access is cheap in cash and expensive in opportunity cost; cap the window and measure the shift in your own data.
• If no perk clears the budget, cut the perk count or raise the threshold — don't quietly eat the margin.

What a per-member perk budget is

A per-member perk budget is the annual dollar cost one qualified member adds for a given benefit, built from your own order data rather than a price list. It answers a different question than eligibility: tier thresholds set with spend percentiles decide who qualifies; costing decides what each qualified member costs you once they do. You need both numbers before you publish a tier list — and if you haven't yet decided whether tiers are worth running at all, the operating signals in when to add loyalty tiers come before any of this.

Build each perk's figure as a formula you can audit: cost per member per year = cost per event × events per member per year. Every perk in this article fits that shape. The events differ — qualifying orders, returns, early-window purchases — but the structure holds, which is what keeps the costing honest when someone argues a perk "doesn't really cost anything."

Common pitfall: costing perks once at launch and never again. Member behavior changes as people climb tiers, so last year's figure silently stops matching this year's cohort.

Free shipping: subsidy per order times qualifying orders

Free shipping costs you the subsidy per qualifying order multiplied by how many orders a tier member actually places. Pull both inputs from the tier cohort itself — the subsidy is your actual carrier cost minus what the member pays, and the order count comes from the qualifying members' real purchase history. Then the arithmetic is one line: if you subsidize $6 per shipment and a qualified member places 8 qualifying orders a year, free shipping costs $6 × 8 = $48 per member per year. That $48 is the number that competes against the tier's margin, not the $9.95 retail shipping rate the member would have paid — and every input here is illustrative, so replace both with your own figures before you budget anything.

A $100 order at a 5% earn rate returns 500 points, redeemed at $0.01 = $5 back — the same discipline applies: work the arithmetic on the page, from your data, not from a published average.

Common pitfall: using retail shipping rates and average-order customers instead of the tier cohort's actual order frequency. Tier members order more often than average — that's why they're in the tier — so an average-based figure understates the cost precisely for the people who get the perk.

Free returns: price the return-rate delta, not the promise

Free returns cost you the change in return rate the perk causes, multiplied by the cost of each incremental return for that tier. The formula: (new return rate − old return rate) × orders × (reverse shipping + restock loss). If free returns lift the cohort's return rate from 10% to 14% — illustrative; measure your own delta — and each incremental return costs $12 in reverse shipping plus restock loss, then 8 orders a year means 0.32 extra returns per member, or roughly $3.84 per member per year before merchandising effects.

Costing the perk at your current return rate gets the answer wrong in a specific direction. The perk's job is to raise the return rate — customers order sizes to try because sending them back is free — so the baseline understates what the perk will actually do. Price the delta, not the promise. Also separate the cash cost from the merchandising effect: an incremental return on an item that sells out anyway costs less than one on an item that misses the season.

Common pitfall: offering free returns tier-wide because it "builds trust," with no return-rate delta measured and no per-member cost on the books. When the delta is unknown, the honest position is that the cost is unknown — run the perk on one tier for a quarter and measure before funding it everywhere. For how returns interact with point balances, the rules in points and returns refund rules apply on top of this costing.

Early access: the perk with a stockout price tag

Early access costs almost nothing in cash and real money in opportunity cost: members buy items during the early window that would otherwise have sold at full price in the public window. The cost per member is the margin you gave away, not the margin you earned — an early-window purchase at full price doesn't cost you anything extra, but a purchase that crowds out a full-price public-window sale does. The practical question is what share of early-window purchases are displaced full-price demand, and only your data can answer it.

Two controls keep this perk affordable. First, cap the window length — 24 to 48 hours, illustrative — so displacement stays bounded. Second, run early access on items with depth, not on your tightest stock, so member purchases don't cannibalize units that would have cleared anyway. Then measure the shift: compare early-window sell-through against comparable launches without early access, from your own data, before you commit to the perk for another year.

The trap: treating early access as a free perk because no invoice arrives. It's the only perk on this list where the cost shows up in the units sold column instead of the expense column, and it's the one most likely to be blamed on "a slow season."

Set the list: perks compete against the tier's incremental margin

Set the tier list by summing each perk's per-member cost, ranking them, and keeping only what the tier's incremental gross margin funds. The threshold comes from your data — incremental margin per member per year, tier cohort only — not from a benchmark, because no benchmark knows your carrier rates, return rates or inventory depth. If the sum exceeds the margin, you have three moves: drop a perk, shrink a perk's scope, or raise the qualification threshold so the cohort's spend funds the list.

Resist copying the big-tier perk menus. Large brands price their perks from a P&L that can absorb a loss-making perk as marketing; you price from yours, and their ability to absorb a perk tells you nothing about whether yours can. A Prime-flavored list on a small shop's margin is how tiers become margin leaks with nice landing pages. The broader cash-and-liability framing lives in building a loyalty program cash and liability budget, and if a perk survives the math only because a supplier pays for it, the funding rules in supplier-funded loyalty perks apply.

Common pitfall: announcing the tier list, then discovering the top tier's perks cost more than the tier earns. The costing work happens before the announcement, every time.

Frequently asked questions

How does costing perks differ from costing points?

Points cost you the redemption value members actually use, and you can model that with earn rates and breakage. Perks cost you per event — per shipment, per return, per early-window order — with no balance and no breakage to soften the bill. Cost both, because a tier that funds its points and not its perks is only half budgeted.

What if the budget funds zero perks?

Then the tier as designed doesn't work. Either the threshold is too low to concentrate enough spend, or the perk set is too rich for the cohort's margin. Raise the threshold, swap cash-cost perks for low-cost ones like early access, or don't launch the tier — an empty tier list costs less than a funded-but-unprofitable one.

How often should I re-cost the perks?

Re-cost when the inputs move: carrier rate changes, a shift in the cohort's order frequency, a return-rate change after a catalog refresh. A yearly re-cost on a fixed calendar is a reasonable default; re-cost immediately after any perk change so the budget never runs on stale inputs.