Why Most Brands Are Losing Money on Amazon in 2026

Key Takeaways

• The average Amazon seller now pays 30–35% of revenue back to Amazon in fees before ad spend or COGS
• FBA fulfillment fees jumped again on January 15, 2026, and a new DD+7 payout delay locked up seller cash starting March 12.
• Average Sponsored Products CPCs hit $1.18 in 2026, up from $0.71 before 2020 — ad cost is the second margin squeeze.
• Most brands are optimizing account-level ACOS while individual SKUs quietly lose money on every sale.
• Only 22% of products on Amazon’s first search page show up in Rufus recommendations — if you’re not optimizing for the AI assistant, you’re invisible to a growing share of shoppers.

Why are so many Amazon brands unprofitable in 2026?

Because two things happened at once, and most brands haven’t recalculated.

Amazon raised FBA fulfillment fees by an average of $0.08 per unit on January 15, and some categories got hit harder — small standard-size items priced $10–$50 saw a $0.25 per unit increase. At 10,000 units a month, that’s $10,000 a year gone, before you’ve paid for a single ad click.

Then Amazon rolled out DD+7 on March 12, which holds seller funds for seven days after delivery confirmation. No opt-out. Legacy sellers with shorter reserve periods lost them.

Meanwhile, third-party seller services — the combined fees Amazon collects from sellers — hit $175.2 billion in 2025, more than Amazon’s own product sales. The marketplace is growing. Your share of that growth isn’t keeping up with what you’re paying to be there.

What does the real math look like per unit?

The account-level numbers lie. The SKU-level numbers tell the truth.

Here’s a standard-size product priced at $29.99, comparing 2024 economics to 2026:

Cost Line	2024	2026	Change
Sale Price	$29.99	$29.99	—
Referral Fee (15%)	$4.50	$4.50	—
FBA Fulfillment	$5.40	$5.65	+$0.25
COGS	$7.00	$7.00	—
PPC Spend (ACOS ~25%)	$5.63	$6.75	+$1.12
Storage + Inbound	$1.20	$1.35	+$0.15
Net Profit per Unit	$6.26	$4.74	−$1.52 (−24%)

Same product. Same price. 24% less profit per unit. (Seller Labs)

When we audit brands at STOCK, we almost always find the same pattern: 10–20% of SKUs are dragging everything else down. The winners cover for the losers, the account-level ACOS looks fine, and nobody notices the negative contribution margins until inventory ties up cash the brand needed for Q4.

What’s not working: optimizing for ACOS instead of profit

Most Amazon ad strategies are still built around ACOS. That’s the mistake.

Average ACOS across Amazon is approaching 30% and climbing — January 2026 hit 32.5%. A 30% ACOS sounds healthy until you realize fees are eating another 30–35% and COGS typically takes another 30–35%. That leaves you fighting over the last 5–10% of revenue, per unit, on every sale.

Here’s what we’re seeing work instead: break-even ACOS at the SKU level. Take your sale price, subtract all non-ad costs, divide what’s left by sale price. That number is the maximum ACOS that SKU can tolerate before ad-driven sales become unprofitable.

Any SKU where actual ACOS exceeds break-even ACOS is losing money on every click — and no amount of bid optimization fixes it. Those SKUs need a price increase, a COGS renegotiation, an FBA-to-3PL switch, or to be pulled from paid media entirely.

How is Amazon Rufus changing the equation?

Rufus is the biggest discovery shift on Amazon since A9, and most brands are not optimizing for it.

Amazon’s AI shopping assistant drove nearly all incremental holiday growth on the platform in late 2025. It’s now generating product recommendations, tracking prices, and answering shopper questions directly in the buying flow.

The problem: Rufus does not rank the same way search does. Research shows only 22% of products on Amazon’s first results page appear in Rufus recommendations, and 36% of Rufus suggestions aren’t on the first results page at all.

If your listings are written for keywords and conversion, they’re underperforming for AI-driven discovery. Rufus pulls from detail page copy, customer questions, reviews, and structured attributes to answer conversational prompts. The brands winning here are rewriting listings around buyer intent and use cases — not keyword density.

This is where LLMO meets retail media, and it’s the opening most brands are missing.

What actually fixes this?

Three moves, in order:

One. Run a full per-SKU profitability audit using 2026 fees. Identify every SKU with a contribution margin under 10% or a negative break-even ACOS gap. Decide which to fix, which to reprice, and which to sunset.

Two. Restructure ad campaigns around break-even ACOS by SKU, not account-level targets. Pull spend from SKUs that can’t support it. Double down on SKUs with margin headroom and strong review velocity.

Three. Rewrite your top 20 listings for Rufus and AI discovery — use-case language, structured attributes, clear value justification at the price point. The goal is to be the product Rufus recommends when a shopper asks “what’s the best X for Y?”

Most brands won’t do any of this because it requires slowing down, auditing the real numbers, and making uncomfortable decisions about which products to stop subsidizing. The ones that do are pulling ahead fast.

---

Want to know which of your SKUs are actually losing money on Amazon right now? We run free per-SKU profitability audits for brands doing $1M+ on the platform. Request yours at stockadmin.co.

Frequently Asked Questions About LLMO

How much do Amazon sellers really pay in fees in 2026?

The average FBA seller pays 30–35% of revenue in combined fees — referral fees, FBA fulfillment, storage, inbound placement, and related charges — before any ad spend or COGS. After ads (typically 20–30% of sales), many brands are left fighting over a 10–15% margin window.

What’s a good ACOS on Amazon in 2026?

There is no universal “good” ACOS. The right benchmark is your break-even ACOS, calculated per SKU based on your actual margin after all 2026 fees. Average account-level ACOS is currently around 30%, but that tells you nothing about whether your specific SKUs are profitable.

How often should I recalculate my Amazon profitability?

At minimum monthly, and immediately after any Amazon fee change. The January 2026 FBA increase alone made previously marginal SKUs unprofitable, and the brands that didn’t recalculate are still running campaigns on math that no longer exists.

What is Amazon Rufus and why does it matter for my listings?

Rufus is Amazon’s AI shopping assistant, now generating a growing share of product discovery on the platform. It pulls recommendations from listing content, reviews, and attributes in ways that differ from keyword search. Brands that don’t optimize listings for conversational, use-case-driven queries are losing visibility to competitors that do.

Should I move off FBA in 2026?

It depends on your unit economics. For slow-moving SKUs or products hit hard by inbound placement and storage fees, a 3PL alternative can restore margin. For fast-moving SKUs where Prime eligibility drives conversion, FBA is still usually the right call. The answer lives in your per-SKU math, not a blanket decision.

---

About STOCK

STOCK is a commerce-first growth agency built for brands that actually have to make the numbers work. We run the unsexy audits most agencies skip — per-SKU profitability, break-even ACOS modeling, listing optimization for both search and AI assistants — and we build the paid media, retail marketplace, and LLMO strategy that makes Amazon a profit engine instead of a treadmill. We work with brands doing $1M+ in annual revenue who are done guessing. stockadmin.co

References

• Amazon Advertising Stats 2026 Update — Ad Badger
• Amazon Seller Profitability in 2026 — Seller Labs
• Amazon Seller Statistics 2026 — Nova Analytics
• Amazon Third-Party Sellers 2025 Q3 & Q4 Data Analysis — AmzMonitor
• While Everyone Debates Agentic Shopping, Amazon’s Rufus is Racking Up Sales — The Drum
• From SEO to AI Agents: Amazon Rufus’s Strategic Shift — Flipflow