Roughly 20% of an Amazon brand's annual revenue runs through four windows: Prime Day in July, Prime Big Deal Days in October, Black Friday/Cyber Monday in November, and the December holiday window.
The steady-state CM1-derived ceiling does not apply during any of them. The standard TACoS ceiling on a 33% CM1 SKU might be 23% in normal weeks. The same SKU might justifiably run at 60% TACoS during Prime Day. Treating the two windows with the same math is one of the most expensive mistakes operators make.
This piece is about how the math actually works during the promo windows that make or break the year.
What breaks during promo periods
Three things break simultaneously, and they compound.
One - CM1 compresses. Coupons stack on top of lightning deal discounts. A 20% Prime Day discount with an additional 10% coupon and a Prime-exclusive 5% can take 35 points off the gross price. CM1 is calculated on the post-discount price, so the dollar margin per unit drops sharply. The 33% CM1 SKU at full price might run at 12-15% CM1 during a stacked Prime Day promo.
Two - CPCs rise. Auction pressure during Prime windows is significant. CPCs commonly rise 30-60% versus normal-week baselines. The cost side of TACoS goes up before any margin compression hits.
Three - the strategic objective changes. During steady-state, the goal is profitable revenue today. During Prime Day, the goal is BSR gain that pays back over the following 60-90 days through improved organic ranking. This is investment math, not contribution math.
The promo period model
The right way to frame promo period economics is as an investment decision, not an operating decision. Four steps.
Step one - pick the strategic objective. Three options:
- Defend share - SKU is at risk of losing BSR rank if it does not participate. The objective is rank-hold; profit is not the metric.
- Take share - Competitor is weak or absent. The objective is rank-grab; the brand is buying organic ranking that will compound for the following quarter.
- Liquidate inventory - Aged inventory needs to clear before storage fees compound. The objective is unit clearance; ad math is secondary to inventory math.
The wrong question is "what is our TACoS target for Prime Day." The right first question is "what are we trying to accomplish during this window."
Step two - model the contribution loss. For each SKU participating, calculate the contribution at the discounted price, ad CPCs at the promo-window rate, and expected unit volume. The total negative or barely-positive contribution for the promo period is the investment cost.
Step three - model the BSR-recovery payback. The objective is to estimate the incremental organic units over the 60-90 days following the promo window, attributable to BSR improvement from the promo-window volume spike. Project organic units at the post-promo BSR level versus organic units at the pre-promo BSR level for each week of the recovery window. Multiply by steady-state CM2 dollars per unit. Sum across 60-90 days.
Step four - approve only if payback exceeds investment by a margin of safety. A 1.5x ratio is a reasonable starting point. If the implied recovery is $150K and the promo-window contribution loss is $100K, the investment passes. If the implied recovery is $90K and the loss is $100K, the investment fails - even though "winning Prime Day" feels strategically necessary.
The discipline most brands miss
Discipline one - pre-stocking ROI is part of the math. Prime Day inventory commitments are made 90-120 days in advance. The working capital tied up in pre-stocking is real, and the BSR-recovery math has to be net of the cost of holding that inventory.
Discipline two - the promo-week ceiling is in dollars, not percent. Percent-of-revenue ceilings break during promo windows because both sides of the ratio move violently. Set the promo-week ceiling in maximum acceptable contribution loss in dollars.
Discipline three - the recovery window is part of the planning, not an afterthought. Most brands plan Prime Day. Few brands plan the 60-90 days after Prime Day.
Per-event variations
Prime Day (July). Highest stakes, biggest BSR impact. Strongest organic-ranking flywheel for the rest of Q3. Heaviest investment justified for SKUs with category-leadership ambitions.
Prime Big Deal Days (October). Smaller volume than Prime Day. BSR recovery overlaps with Q4 build-up, so payback math has to weight Q4 organic differently than the Q3 baseline. Brands often under-invest here relative to the actual opportunity.
Black Friday / Cyber Monday (November). Highest CPC inflation of any window. Margin compression is severe. The brands that win BFCM are usually the brands that pre-built Q3 BSR through Prime Day and ride the higher organic position into the cheaper organic Cyber Week traffic.
December holiday window (mid-December). Often the most profitable of the four for Amazon-native brands because gift-buying drives high-intent traffic. CPC inflation is real but not as extreme as BFCM.
When the right answer is to skip the promo window
Case one - high-margin niche with no BSR competition. If the SKU is dominant in a niche with no real BSR-threatening competition, the promo-week investment does not pay back.
Case two - supply-constrained brands. If inventory is the binding constraint year-round, participating in promo windows accelerates stockout. Better to skip Prime Day, hold the inventory, and capture the higher steady-state margin across more weeks of the year.
Case three - subscription or repeat-purchase brands. Brands whose LTV is heavily back-loaded may find that the Prime Day customer is the worst-acquired customer of the year (discount-driven, low repeat probability). Math the discounted-customer LTV against the BSR-recovery payback before participating.
How this integrates with the rest of the model
Integration with unit-economics ceiling. Promo periods are explicit, planned, time-boxed breaches of the steady-state ceiling. They are not exceptions to the model - they are a different model for ~20% of the year.
Integration with working capital ceiling. Promo windows accelerate working capital consumption. Pre-stocking commits cash 90-120 days early. Brands operating near their working capital ceiling cannot fully participate in promo windows without either external capital or a deliberate slowdown elsewhere.
Integration with the portfolio framework. Not every SKU should participate in every promo window. High-revenue × high-CM1 SKUs justify promo investment because the BSR-recovery math compounds; low-revenue × low-CM1 SKUs should usually stay out because the math does not pay back.
Bottom line
The steady-state ceiling does not apply during promo windows. Treating Prime Day with the same math you use in March will either price you out of participating or break your margin defending what you should not have been defending in that window.
The right model for promo periods is investment math: maximum acceptable contribution loss in dollars, time-boxed to the promo week, justified by implied BSR-recovery payback over the following 60-90 days, approved only if recovery exceeds loss by a margin of safety.
If you want a second read on your promo-window investment math for the coming year, the Amazon Account Audit covers exactly this work.
Frequently asked questions
Why doesn't the standard TACoS ceiling apply during Prime Day?
Three things break simultaneously during promo windows: CM1 compresses from coupon and discount stacking, CPCs rise materially from auction pressure, and the strategic objective shifts from contribution to BSR-recovery investment.
How should I think about Prime Day budget?
As an investment, not as a TACoS target. Calculate the contribution loss for the promo week. Estimate the BSR-recovery payback over the following 60-90 days. Approve the spend only if recovery exceeds loss by a margin of safety.
Should every SKU participate in every promo window?
No. High-revenue and high-CM1 SKUs justify promo investment because BSR-recovery compounds. Low-revenue and low-CM1 SKUs usually should not participate. Supply-constrained brands and subscription brands have specific reasons to skip certain windows entirely.
Why should the promo-week ceiling be in dollars instead of percent?
Percent-of-revenue ceilings break during promo windows because both sides of the ratio move violently. The dollar floor for maximum acceptable contribution loss is more honest and easier to defend than any percent-based ceiling during promo windows.
What's the most common promo-period mistake?
Skipping the 60-90 day recovery planning. Most brands plan Prime Day. Few brands plan the operating posture for the 60-90 days afterward.