How to Set Your Peak-Season Floor Price and Discount Depth

Your Floor Price Is Probably 25% Too Low

Most stores set a peak-season floor the same way they set it in March. Take landed cost, add shipping, round up, and call that the number you will not go below. On a product that costs $22.00 to land and $7.40 to ship at peak rates, that method produces a floor of $29.40.

The real break-even on that product is $36.82. Every discount you plan sits on top of a number that is 25% too low.

Nothing exotic causes the gap. Returns strip revenue you already booked while leaving most of the fulfilment cost behind. Peak carrier surcharges are published in advance and rarely make it into the floor. Payment processing is charged on the gross sale and is not handed back when the sale reverses. These are ordinary costs. They are simply left out of the one number that decides how deep you are allowed to cut.

This post is the arithmetic, and the arithmetic is the whole point. Two formulas do the work: the floor price with returns and peak shipping, and the break-even volume for a given discount. Both are written so you can substitute your own figures, which matters, because the numbers that decide this are yours and not your category's.

What Actually Moves in Your Costs Between September and December

Three cost lines shift at peak, and only one of them arrives as an invoice line you can read.

Carrier surcharges are the knowable one. FedEx's 2026 peak schedule adds a Ground Residential demand surcharge of $0.50 to $0.80 per package, with surcharges starting on September 28, 2026 (Supply Chain Dive, July 2026). That is a published number you can put straight into the formula in September. UPS has not announced its equivalent schedule at the time of writing, so hold a placeholder at a similar magnitude and re-check in October. The peak-season pricing checklist covers the windows and dates in detail, so treat the figure here purely as an input.

Returns are the expensive one, and most stores account for them as a logistics cost rather than a pricing input. Retailers surveyed by NRF and Happy Returns expected 17% of holiday sales to be returned, against an online return rate of 19.3% across the year (NRF, October 2025). That survey covered merchants above $500 million, so use 17% as a planning placeholder only until your own January reconciliation gives you a real rate. Your own number is the one that belongs in the formula.

Payment processing is the one nobody can size for you. Your contractual rate and fixed fee are known, but the Q4 mix shifts between card types, wallets and buy-now-pay-later in ways that move the blended rate. No tier 1 to 3 source quantifies that lift, so treat it as a sensitivity you test at plus or minus a few tenths of a percentage point, not as a line item you can pin down.

Two other gaps are worth naming plainly rather than papering over. There is no current, verifiable US retail gross-margin benchmark to compare yourself against, and there is no credible 2025 to 2026 study quantifying promotional lift against margin erosion. Where this post needs a margin figure, it uses yours. Any post that hands you a tidy table of category margin bands is inventing it.

Formula One: The Floor Price That Survives Returns

Start with expected contribution per shipped unit, because a unit that ships is not the same as a unit that stays sold.

E = (1-r)·P·(1-m) - (1-r)·f - [ (1-r)C + r·C(1-k) ] - S - F - r·R

Every term, per shipped unit:

• P is the price you charge.
• C is landed unit cost.
• S is outbound shipping and packaging at peak rates, surcharges included.
• F is pick and pack.
• R is return handling plus inbound freight, per returned unit.
• r is your return rate, as a fraction.
• k is the fraction of a returned unit's cost you recover by reselling it at full value.
• m is your processing rate, as a fraction.
• f is the fixed processing fee per transaction.

Read the expression left to right. You keep the price only on units that are not returned, and only after processing takes its percentage, hence (1-r)·P·(1-m). The fixed fee applies to the same surviving transactions. The bracketed term is the cost of goods: units that stay sold cost you C, and units that come back cost you C(1-k), because resale recovers the fraction k. Shipping, pick and pack are spent on every unit that leaves, returned or not, so they carry no (1-r). Return handling is spent only on the fraction that comes back.

Because (1-r) + r(1-k) simplifies to 1 - r·k, setting E = 0 and solving for P gives the floor:

                C(1 - r·k) + S + F + r·R + (1-r)·f
P_floor  =  ---------------------------------------
                      (1 - r)(1 - m)

The structure is what matters more than the symbols. The numerator is everything you spend per shipped unit. The denominator is the fraction of the sticker price you actually keep. Cost plus shipping is what you get if you pretend that denominator equals 1.

The Worked Example

Take C = $22.00, S = $7.40 at peak rates, F = $2.00, R = $6.00, r = 0.17, k = 0.80, m = 0.029, f = $0.30.

The numerator, term by term:

• C(1 - r·k) = 22.00 × (1 - 0.136) = 19.008
• S = 7.40
• F = 2.00
• r·R = 0.17 × 6.00 = 1.02
• (1-r)·f = 0.83 × 0.30 = 0.249

That totals 29.677. The denominator is 0.83 × 0.971 = 0.80593. Divide, and P_floor = $36.82.

Cost plus shipping said $29.40. The true break-even is 25% higher. Sell that product at $32 during a flash sale, feeling comfortable because you are still above your floor, and you lose money on every unit while the dashboard shows revenue climbing.

Adding a Target Contribution

A break-even price is not a business. Zero contribution means the product pays for itself and nothing toward overheads, marketing or the ad spend that drove the sale.

To hold a target contribution rate c as a fraction of price, subtract c from the whole denominator: (1 - r)(1 - m) - c. Note where the c sits. Putting it inside the bracket would define your target against post-returns revenue instead of against price, and the second formula below reads c as a fraction of list price, so the two would stop agreeing. At a 15% target the denominator is 0.80593 - 0.15 = 0.65593, and the same numerator of 29.677 gives $45.24.

Two numbers now bracket the product. Below $36.82 it destroys value. Below $45.24 it fails to fund the rest of the business. Between them you are choosing how much of the overhead contribution to give away in exchange for volume, which is a decision, not an accident.

Formula Two: How Much Volume a Discount Has to Buy

Discounting is a bet that lower price buys more units than it costs in margin. The break-even for that bet is one line of algebra.

With contribution rate c and discount depth d, both as fractions of list price:

V1/V0 = c / (c - d)        required unit uplift = d / (c - d)

The derivation takes one step. Contribution per unit at list price is cP. A discount of depth d cuts the price by dP and leaves variable cost unchanged, so contribution per unit becomes P(c - d). To hold total contribution flat, units must rise by exactly the ratio of the two, which is c / (c - d).

What Category Depth Actually Demands

Salesforce's holiday forecast put the average US discount rate at 29% during Cyber Week 2025 and around 23% across the full November 1 to December 31 season, with category depth ranging from 37% in general apparel and 35% in health and beauty down to 23% in home (Salesforce Shopping Index, via Digital Commerce 360, September 2025). The same forecast projected $288 billion in US online holiday spend, up 2.1% against 4% growth the year before, and described retailers as more cautious with promotions (Salesforce, via Investing.com, September 2025).

Every figure in that paragraph comes from a single upstream, the Salesforce Shopping Index, and every one is a forecast published in September 2025 rather than a measurement taken after the season. Treat them as a planning assumption about where your category may land, not as evidence of where it did.

Run those depths through the formula, against illustrative contribution rates:

Category depth	c	d	Units needed
Apparel	0.55	0.37	x3.06 (+206%)
Health and beauty	0.60	0.35	x2.40 (+140%)
Home	0.35	0.23	x2.92 (+192%)
Cyber Week average	0.40	0.29	x3.64 (+264%)

An apparel seller at a 55% contribution rate discounting 37% needs to triple units to stand still. At the Cyber Week average of 29% off against a 40% contribution rate, the promotion needs 3.64 times the baseline volume, an uplift of 264%, before it has earned back what it gave away. Very few peak promotions do that. Most are judged on revenue growth, which rises comfortably at 1.5x volume while contribution falls through the floor.

The contribution rate you feed in has to be the peak rate from the first formula, after returns and surcharges, not the gross margin on your price list. Using list gross margin here is the single most common way this calculation flatters a promotion that is losing money. If you want the demand side of the same question, price elasticity is what determines whether the required uplift is even reachable.

The Point Where No Volume Saves You

There is a hard limit in the formula, and it is worth stating on its own.

When d reaches c, the denominator (c - d) goes to zero and break-even volume is infinite. Every unit contributes exactly nothing, so selling ten times as many changes nothing. Past that point, when d exceeds c, the ratio turns negative, which is the algebra's way of saying the promotion cannot break even at any volume: each additional unit deepens the loss.

No merchandising cleverness moves this. It is not a threshold you can push with better bundling or a stronger email campaign. Above your contribution rate, volume is the enemy.

The Verdict on the Worked Example

Put the two formulas together on the product from earlier, listed at $59.99.

Its true break-even is $36.82, which is 1 - 36.82/59.99, or 38.6% off. Discount deeper than that and every unit loses money.

Its 15% contribution price is $45.24, which is 1 - 45.24/59.99, or 24.6% off. That is the deepest cut the product can carry while still funding overheads.

The forecast Cyber Week average is 29% off. This product cannot reach it. Not "should not", cannot: matching the category average puts it 4.4 percentage points past its own target-contribution limit, and the seller who matches it has quietly agreed to fund the difference out of overhead.

That is the moment worth sitting with. The category average is a description of other people's cost structures, aggregated. Their landed costs, return rates and shipping profiles are not yours. When your arithmetic says 24.6% and the category says 29%, the category is not a target you are failing to hit. It is a number that does not apply to you.

Tier by Contribution Rate, Not Revenue Rank

Most catalogue tiering sorts products by revenue: hero, mid, long tail. That framing is covered in the competitive pricing strategy guide and it works well for deciding where to spend monitoring attention.

For discount depth it is the wrong sort key. Sort by peak contribution rate instead. A high-revenue product with a thin post-returns contribution rate has less discount headroom than a mid-tier product with a fat one, and revenue rank hides exactly that. Run the floor formula across your catalogue, sort descending by the resulting contribution rate, and set maximum depth per band. The products at the bottom of that list are the ones that must never be handed a category-average discount.

Where Peak-Season Monitoring Fits

Your maximum depth is fixed by arithmetic, but whether it is competitive is an empirical question about what other sellers do. That is worth knowing before you commit, not after. The Black Friday price monitoring playbook covers the phased approach to building that baseline and why September is when it has to start. The short version: a watch list built in November records already-discounted prices as normal, which makes every later comparison meaningless.

Respot handles the recording side. You paste a competitor's product URL, it detects the product details automatically, and it tracks price and stock per variation rather than at parent-product level, which matters when only one size or color competes with your bestseller. The free plan gives you 5 trackers with a 7-day price history and no credit card, enough to test whether your contested products behave the way this arithmetic predicts. Paid plans run to 100, 400 or 2,000 trackers with longer history as the list grows.

If your arithmetic caps you below the category, the monitoring question changes shape. You are no longer looking for a price to match. You are looking for the moments when competitors sell out, because that is when demand moves to you at a price you can actually afford to hold.

Do the Arithmetic Before You Commit the Depth

The decision this post asks you to make is small and unglamorous: calculate your real floor before you publish a single peak-season discount, and calculate it per product rather than per category.

It takes eight inputs, most of which you already have. Landed cost and processing terms come off your contracts, shipping at peak rates is published, and the return rate is the one you estimate this year and measure next January. Twenty minutes with a spreadsheet gives you a floor and a maximum depth per product, and those two numbers turn every promotional decision from a feeling into a check.

The sellers who come out of peak season with margin intact are rarely the ones who discounted least. They are the ones who knew, per product, exactly how far down they could go, and stopped there while everyone else followed the category average off a cliff. Start tracking what your competitors actually charge while there is still time to build a baseline.