Amazon Selling Guide 2026: AI Search, Fees and Q4 Strategy

Amazon retired the Rufus name on a Wednesday in May.

Most sellers I talk to still haven't noticed. They're writing listings for a keyword box that isn't the front door anymore, paying a fuel surcharge they never budgeted for, and planning Q4 off last year's calendar.

So here's the short version of this Amazon selling guide 2026. Shoppers now ask Amazon's AI assistant, Alexa for Shopping, what to buy, and it answers from your listing, your reviews and your Q&A, so every title, bullet and image has to answer a real question. Your costs went up twice: a 3.5% fuel and logistics surcharge on fulfillment fees since April 17, and holiday peak fees averaging $0.32 per unit from October 15, 2026 to January 14, 2027. Your Black Friday stock has to land between October 14 and October 28, depending on how you ship. Every deal and every ad dollar has to clear a break-even number you calculate before you spend, not after. And your best product deserves a store you own.

That's the whole game. The rest of this page is the how, with the math and the sources.

On Amazon you're a tenant. The rent went up this year, and the landlord hired an AI to decide which shelf you stand on.

I built brands on Amazon before I built RevenueFlows AI. I'm not anti-Amazon. It's still the biggest shelf in the world. But I've watched too many good products earn a few dollars a sale for years because the seller never ran the numbers on anything else. This guide is the one I'd hand a friend who sells on Amazon today.

How Amazon's AI shopping assistant decides what gets recommended

Start with what actually changed.

On May 13, 2026, Amazon unified Rufus and Alexa+ into one assistant called Alexa for Shopping. It isn't a side chat window anymore. According to Amazon's own announcement, shoppers can ask questions right in the main search bar, get AI overviews in search results and on product pages, build shopping guides for big purchases, compare products side by side, see up to a full year of price history, and set up automated deal alerts and routine purchases. It's free for every US customer on the app and website, no Prime membership required. Amazon says Rufus alone helped over 300 million customers in 2025.

So how does it decide which product to recommend?

Here's the honest answer: Amazon hasn't published a ranking formula, and anyone who tells you they've reverse-engineered it is selling something. What Amazon has published is the raw material. When it launched Rufus, Amazon said the assistant was trained on its product catalog, customer reviews, community Q&As and information from across the web. Alexa for Shopping keeps that product knowledge and adds each shopper's own history and preferences.

Read that list again. Catalog, reviews, Q&A, the web.

That's your listing, what buyers say about it, what buyers asked about it, and what the rest of the internet says about it. Three of those four you can influence this week.

What this means in plain words

The old game was keyword matching. A shopper typed "bamboo sheets queen" and the algorithm looked for listings that contained those words and sold well.

The new game is question answering. A shopper types "which sheets stay cool for a hot sleeper who hates pilling" and an AI reads through listings and reviews looking for an answer. If your listing says "400 thread count, 100% viscose from bamboo, deep pocket" and nothing else, the AI has specs but no answer. If your bullets say who the sheet is for, what problem it solves and what it won't do, the AI has something to quote.

The keyword box rewarded sellers who stuffed words. The AI rewards sellers who answer questions.

My take, and I'll own it as an opinion: this is the best thing to happen to good products on Amazon in years. For a decade, the listing that won was often the one with the most keywords crammed into a 200-character title. Now the listing that wins is more likely to be the one that sounds like the best salesperson in the category.

Two features deserve extra attention.

Price history. Shoppers can now see up to a year of price history. If you inflate a list price before a deal so the discount looks bigger, the shopper can see it. Don't play that game in 2026.

Deal alerts. Amazon's Prime Big Deal Days announcement (the event runs October 6 and 7) tells shoppers to set alerts like "find me deals on espresso machines." That means your deal can find a buyer who never searched for you, as long as the AI understands what your product is and who it's for.

Want to test your listing against the questions shoppers ask? Our free listing checker built for Amazon's AI assistant (named before the rebrand, same job) shows which common buyer questions your copy answers and which it dodges.

Listings that answer questions: titles, bullets, A+

Here's the part most sellers are about to get wrong.

On June 10, Amazon posted an announcement in Seller Central: starting July 27, 2026, titles in all categories except media must be 75 characters or less, including spaces. Amazon added a new field called Item Highlights with 125 more characters for materials and recommended uses, and says that text is searchable and shows with titles in search results and on detail pages.

And here's the line nobody should skim past. Amazon says any titles still over 75 characters after July 27 "will be updated to the AI recommendation gradually," and brand owners get 14 days to review, change or approve those AI-written titles and Item Highlights.

Read it this way: if you didn't rewrite your title, Amazon's AI will write one for you. Maybe it'll be good. Maybe it'll drop the one word that separates you from the knockoff two listings down. Either way, you're letting the landlord repaint your storefront.

A 200-character title was a keyword dump. A 75-character title is a promise. Write the promise yourself.

Titles: 75 characters, one promise

With 75 characters you get brand, product, the one attribute that decides the sale, and maybe a size or count. That's it. So pick the attribute that answers the buyer's first question, not the one your supplier listed first.

Then move the second-tier facts into Item Highlights. Materials. Use cases. Who it's for. Since Amazon says Item Highlights is searchable, it now carries some of the weight your old title carried.

If you want a second opinion, our free Amazon title tool checks your title against the new limit and flags what's wasting space.

Bullets: objections first, specs second

Most bullets I audit read like a spec sheet. Five lines, each starting with a capitalized feature, none answering the question that's actually stopping the buyer.

I've written about this before. In our Amazon bullet points study we compared feature-first bullets against objection-first bullets, and in 50 Amazon listing audits the same flaw kept showing up: listings that describe the product instead of answering the buyer. The fix is the same one that works for the AI.

Here's a simple test. Take each bullet and ask: what question does this answer, in the buyer's words?

• "Premium 18/8 stainless steel" answers nothing.
• "Won't rust or hold coffee smell after a year of daily use" answers "will this thing get gross?"

The second one is also the kind of sentence an AI assistant can lift straight into an answer.

Go through your last 100 reviews and your Q&A. Write down every question buyers asked more than twice. Those are your bullets. If a question shows up in 1-star reviews, it goes in bullet one.

A+ content: the comparison and the "who it's for"

A+ is where you get room. Don't waste it on another lifestyle banner with three words on it.

I treat A+ as the place for two things the bullets can't fit: a comparison chart (your sizes, your bundle options, or you against the generic alternative, described honestly) and a clear "who this is for, and who it isn't for" section. Buyers trust a brand that tells them when not to buy. It also cuts returns.

If you want a score on your current modules, the free A+ content grader grades each one. And if you'd rather have a person look at the whole listing, our breakdown of what an Amazon listing audit service should find covers what to ask for.

Images and AI image rules

Images are where AI tools save sellers the most time, and where sellers get the most confused about the rules.

Here's what Amazon has actually said. In a Seller Central reply from an Amazon staff member, Amazon laid out four points:

1. Simple AI retouching, like background removal, color correction and lighting fixes, is generally treated the same as traditional photo editing.
2. Images with photorealistic AI-generated people need to be disclosed, and A+ uploads now have checkboxes for "AI-generated" and "AI-generated people."
3. You don't have to go back and label content that's already published. The disclosure is required for new uploads going forward.
4. The main image still needs a pure white background, the product filling at least 85% of the frame, and at least 500 pixels on the longest side, with 1,600 or more recommended so zoom works.

That's the rulebook. Now the strategy.

AI can make you a beautiful image in 30 seconds. It can't make the buyer believe it. Your images still have to be your product.

Main image: real product photo, clean white background, retouched as much as you like. Don't use a fully synthetic render as the hero. The main image is a promise about what shows up in the box, and a render that looks better than the real thing is a return waiting to happen.

Secondary images: this is where AI earns its keep. Lifestyle scenes, scale shots, infographics that answer the top three questions from your reviews. If a scene includes a photorealistic AI-generated person, disclose it when you upload.

The image stack I'd build for most products: main image, a scale shot next to a common object, a "what's in the box" shot, an infographic answering the number one objection, a lifestyle image showing the moment of use, a comparison image, and a short video if your category supports it.

Before you upload anything AI-assisted, run it through our free AI image policy checker. Then grade the whole gallery with the listing images grader to see which of those seven slots you're missing.

2026 fee changes and Q4 peak fees

Now the money. This is where a lot of Amazon sellers are quietly losing margin they don't know they've lost.

There were three fee moves this year that matter for Q4.

January 15, 2026: the annual fee update. Amazon's 2026 fee announcement said FBA fees would rise by an average of $0.08 per unit sold. Small on its own.

April 17, 2026: the 3.5% fuel and logistics surcharge. Amazon added a 3.5% surcharge on fulfillment fees for FBA in the US and Canada from April 17, extending to Buy with Prime and Multi-Channel Fulfillment on May 2. It applies to the fulfillment fee, not your sale price, and trade press reporting puts it at about 17 cents per unit on average for US FBA, depending on size. Amazon called it temporary but gave no end date.

October 15, 2026 to January 14, 2027: holiday peak fulfillment fees. Amazon's holiday 2026 announcement puts the peak increase at an average of $0.32 per unit, and trade press reporting confirms the 3.5% surcharge stacks on top. The peak fees apply to FBA, Remote Fulfillment with FBA, Multi-Channel Fulfillment and Buy with Prime.

That last point matters if you already use Multi-Channel Fulfillment to ship your Shopify orders from Amazon's warehouses. Your own-store orders pay the peak fee too.

And underneath all of it sits the fee most sellers stopped seeing years ago: the referral fee. Amazon's own pricing page lists 15% for most categories. Our Amazon seller page puts it bluntly: "Platforms like Amazon eat 15% of every sale, which means you're paying $150k in referral fees for every $1 million in sales."

Here's every date in one place.

Date (2026)	What happens	Source
January 15	FBA fees rise an average of $0.08 per unit	Amazon fee announcement
April 17	3.5% fuel and logistics surcharge on FBA fulfillment fees (US, Canada)	Trade press report
May 2	Surcharge extends to Buy with Prime and Multi-Channel Fulfillment	Trade press report
May 13	Rufus folded into Alexa for Shopping	GeekWire
July 27	Titles capped at 75 characters (not media); AI rewrites long titles	Seller Central
October 6 to 7	Prime Big Deal Days	About Amazon
October 14	AWD inventory arrival for Black Friday and Cyber Monday	Seller Central
October 15	Holiday peak fulfillment fees start (average $0.32 per unit)	Seller Central
October 20	Last day to submit Black Friday Week and Cyber Monday deals	Seller Central
October 21	FBA arrival, minimal shipment splits	Seller Central
October 28	FBA arrival, Amazon-optimized shipment splits	Seller Central
October 31	Last month of off-peak storage rate for AWD auto-replenishment	Seller Central
January 14, 2027	Holiday peak fulfillment fees end	Seller Central

A word of caution on the averages. "$0.32 per unit" and "17 cents per unit" are averages across everything Amazon ships. Your number depends on your size tier and weight, so pull your own fee preview in Seller Central. If you want the full stack on one screen, our free FBA fee calculator lays out referral, fulfillment and storage per unit.

The fee that hurts you isn't the one you see on the invoice. It's the 30 cents a unit you never put back into your price.

Inventory deadlines, restock limits and storage fees

Today is the last week of September. If you sell anything giftable, the next 30 days decide your Q4.

Here's the calendar that matters, straight from Amazon's holiday announcement. For Black Friday Week and Cyber Monday, inventory needs to arrive by:

• October 14 if it's coming through Amazon Warehousing and Distribution (AWD)
• October 21 for FBA shipments with minimal shipment splits
• October 28 for FBA shipments using Amazon-optimized splits

Those are arrival dates. Your freight has to be moving well before them.

Two warnings from the same period.

First, capacity. Trade press reporting on the holiday fees notes that sellers may run into lower capacity limits during the holiday rush, as fulfillment centers shift focus to processing customer orders in November and December. So the restock limit you had in August isn't a promise for November. If your limits tighten, AWD or a 3PL can hold the overflow, but only if you set it up before the rush.

Second, storage. Amazon's own FBA fee guide says storage costs vary by season, with higher rates in the holiday months. The trap runs both ways. Send too little and you stock out on the best selling days of the year and lose ranking you spent months building. Send too much and you pay holiday storage rates on units that sit into January.

Here's how I think about the number. Take your daily sales rate from last November and December (or your current rate times the seasonal lift you actually saw last year, not the one you hope for). Multiply by the days between your arrival date and the date you can realistically restock in January. Add a safety buffer for your lead time. Then compare the storage cost of the buffer against the profit on the sales you'd lose to a stockout. Most of the time, a few weeks of buffer is cheaper than a stockout on a product that ranks.

And once you have a quantity, run it through the storage fee forecaster to see the storage bill before it lands.

A stockout in December costs you the sale. It also costs you the ranking that would have sold the next hundred.

One more thing that sellers forget: the calendar also affects your listing. Any listing change you want live for Black Friday traffic should be done before the inventory ships, not the week of. Black Friday shoppers are the coldest, most price-checking buyers of the year. We wrote a whole piece on writing a product page for Black Friday traffic for Shopify, and most of it applies to your Amazon listing too.

Deals and coupons: the profit math

This is where I get the most pushback, so let me show the math instead of arguing.

First, the fees. Amazon's holiday announcement says event deal fees are unchanged from Prime Day: an upfront fee of $100 per promotion plus 1.5% of promotional sales, capped at $5,000. Coupons run on the structure Amazon introduced in June 2025, an up-front fee of $5 per coupon plus 2.5% of coupon sales. Check both in Seller Central before you commit, since promotion fees get revised.

Now run the math on a product like this (a hypothetical, not a client).

A $40 product. After the 15% referral fee, fulfillment, landed cost and normal ad spend, it keeps $9 profit per unit. You sell 500 units a month at full price. That's $4,500 profit.

Now you run a 20% off deal. The price drops to $32. You lose $8 of revenue per unit. Assuming the referral fee is charged on the $32 the buyer actually pays, it drops from $6.00 to $4.80, so you get $1.20 of that back. Your profit per unit is now $9 minus $8 plus $1.20, which is $2.20. Then the 1.5% variable promotion fee takes $0.48 per unit. You're at $1.72.

Here's the math that matters. To make the same $4,500 you'd make at full price, plus the $100 upfront fee, you need $4,600 divided by $1.72. That's about 2,675 units. More than five times your normal month.

Watch what happens to the per-unit profit.

Full price $9

20% deal before promotion fee $2.20

20% deal after promotion fee $1.72

Hypothetical: $9 profit at full price. 20% off removes $8, the referral fee drops $1.20, the 1.5% promotion fee takes $0.48.

Can a deal sell five times your normal volume? Sometimes. A strong Prime Big Deal Days slot or Black Friday placement can. And a deal can buy ranking that pays off for months. But I want you to make that bet on purpose, with the number in front of you, instead of running a deal because everyone else is.

A discount doesn't fix a listing. It pays the buyer to overlook what the listing never said.

Three rules I'd follow:

1. Deal on products with fat margins, not thin ones. On a product keeping $4 a unit, a 20% discount on a $40 price wipes you out before fees.
2. Deal to rank, then stop. If the deal gets you to page one for a term that converts, the payoff is the full-price sales after the event.
3. Never inflate the list price first. With a year of price history in front of shoppers, a fake strikethrough is easy to spot.

Before you set up a coupon, the free coupon profitability calculator shows your net per sale after the coupon and its fees.

PPC: audits, automation and break-even advertising cost of sales

Advertising cost of sales (ACoS) is the one PPC number every Amazon seller knows, and the one most sellers read wrong.

Amazon's own ACoS guide defines it as ad spend divided by ad revenue, times 100. Spend $50, sell $100, your ACoS is 50%. And the same guide says the thing that matters: to stay profitable, your ACoS needs to be lower than your profit margin. Otherwise you're spending more on ads than you're earning.

So your target isn't an industry benchmark. It's your break-even ACoS, and it's different for every product.

The break-even math

Run the numbers on a product like this (hypothetical again).

A $40 product keeps $12 per sale after the referral fee, fulfillment and landed cost, before any ad spend. $12 divided by $40 is 30%. That's your break-even ACoS. Every ad sale at 30% ACoS makes you exactly $0.

Want profit on ad sales? Aim below it. At 20% ACoS you spend $8 per $40 sale and keep $4.

Now add Q4. If this product's peak fulfillment fee matches Amazon's $0.32 average, your pre-ad profit drops to $11.68, and your break-even ACoS drops to 29.2%. Small change. But it's across every unit from October 15 to January 14, and most sellers never update the number.

Automation: let Amazon bid, but set the guardrails

Amazon Ads now does a lot of the bid work for you. The one to understand first is dynamic bidding. With dynamic bids, up and down, Amazon raises or lowers your bid in real time based on how likely the click is to convert, and it can raise a bid by up to 100%. A $1.00 bid can become $2.00. For Q4, Amazon also offers event-based bid rules that raise bids automatically during high-traffic events.

Automation is good at one thing: finding the auction you can win. It doesn't know your margin. It will happily buy you sales at a 45% ACoS on a product that breaks even at 30%.

So here's how I'd split the work:

• You set the ceiling. Break-even ACoS per product, written down.
• The machine bids inside it. Dynamic bids and event rules for the peak days.
• You audit every two weeks. Pull the search term report. Cut terms that spend with no sales. Raise bids on terms converting under your break-even.

That audit is where most of the money hides. If you've never done one, our free Amazon PPC audit tool flags the usual leaks: spend on terms with zero orders, campaigns running over break-even, and broad-match terms eating budget.

Automation is a great employee and a terrible boss. Give it a margin to work inside.

One last point on ads, and it's the one that matters most. The cheapest way to lower your ACoS is a listing that converts better. Same clicks, more sales, lower ACoS. That's why the listing work at the top of this guide comes before the PPC work. Our piece on Amazon product page conversion goes deeper on the listing side.

TikTok Shop and Walmart as second channels

Every seller who gets squeezed on Amazon asks the same question: where else should I sell?

The two names that come up most are TikTok Shop and Walmart Marketplace. Both can work. Both are also someone else's shelf. Keep that in mind.

Walmart Marketplace is the closest thing to Amazon without being Amazon. Walmart's pricing page says there are zero setup or monthly fees, and referral fees range from 3% to 20% depending on category. It also runs its own fulfillment service, WFS. Walmart suits products that already win on price and reviews, since the buyer behavior is close to Amazon's: search, compare, buy. Run your numbers in the free Walmart fee calculator before you list.

TikTok Shop is a different animal. On Amazon and Walmart, the buyer is searching. On TikTok, the buyer is scrolling, and they weren't planning to buy anything. That means demonstrable products win: things you can show working in the first 3 seconds. It also means the listing has to sell to someone who arrived cold. We covered that exact problem in how to write a product page for TikTok traffic.

On fees, I'll be straight with you: 2026 sources don't agree on TikTok Shop's current referral rates by category, so pull your rate from your own Seller Center and plug it into the TikTok Shop fee calculator before you model anything. Don't build a plan on a blog post's number, including mine.

My opinion on sequencing: add one second channel, not three. Pick the one that matches how your product sells. Search-driven products go to Walmart. Show-driven products go to TikTok. Then give it 90 days before you judge it.

But here's the thing. Walmart and TikTok spread your risk across more landlords. They don't make you an owner. For that, you need a store.

When and how to add your own store

This is the section I care about most, because it's the one that changed my own thinking years ago.

On Amazon, the buyer is Amazon's customer. You don't get their email. You can't send them a new product. And you pay 15% of every sale for the privilege, before fulfillment, storage, ads and now a fuel surcharge. On your own store, the buyer is yours, and so is the margin.

Here's what that looks like with real numbers. These are three client products on our results page, each moved from Amazon FBA to its own Shopify store. Real client numbers, not typical results, and not a promise of what your store will do.

Product 1: "$8K/mo on Amazon to $30K/mo on Shopify." "Instead of making $17 per order, we started banking a fat $48 profit per order, after paying for ads and inventory."

Product 2: "Took a $30 Amazon product to $177 per order on Shopify." "Instead of making $30 per order, we started banking a fat $78 profit per order, after paying for ads and inventory."

Product 3: "Brand-new product to $50K/mo in 6 months." "Instead of making $50 per order, we started banking a fat $134 profit per order, after paying for ads and inventory."

Here are the full breakdowns, straight from the results page.

	Product 1 Amazon	Product 1 Shopify	Product 2 Amazon	Product 2 Shopify	Product 3 Amazon	Product 3 Shopify
Product price	$17.95	$24.95	$29.95	$49.95	$49.95	$79.95
Platform fees	-$2.69 (15%)	-$0.75 (3%)	-$4.49 (15%)	-$1.50 (3%)	-$7.49 (15%)	-$2.40 (3%)
Gross profit per sale	$6.52	$13.40	$13.72	$34.65	$27.57	$60.20
Advertising costs	-$2.69	-$26.36	-$5.99	-$63.40	-$9.99	-$43.02
Average order value	$17.95	$101.38	$29.95	$177.75	$49.95	$210.86
Net profit per sale	$3.83	$48.27	$7.73	$78.87	$17.58	$134.28

Product 1 on Amazon $3.83

Product 1 on Shopify $48.27

Product 2 on Amazon $7.73

Product 2 on Shopify $78.87

Product 3 on Amazon $17.58

Product 3 on Shopify $134.28

After paying for ads and inventory. From the three Amazon-to-Shopify case studies on revenueflows.ai/results.

Look at the advertising line. The Shopify side spends far more on ads per sale. Product 1 spends $26.36 per sale on Shopify versus $2.69 on Amazon. And it still makes more than 12 times the net profit per sale ($48.27 against $3.83).

Why? Look at the average order value line. On Amazon, the buyer adds one unit to a cart full of other people's products. On your own store, the page can sell the bundle, the two-pack, the refill, the upgrade. Product 1's average order value went from $17.95 to $101.38. Same product. Different shelf.

On Amazon you compete for the sale. On your own store you design the sale.

When to make the move

Not every product should leave. Here's my rule of thumb:

• Move a product that already sells. Amazon is a great place to prove demand. Once a product has steady sales and reviews that say people love it, it's ready.
• Move a product that bundles. If buyers naturally want two, three or a refill, your store can sell that. Amazon's cart mostly can't.
• Keep Amazon running. This isn't a breakup. Amazon stays your search channel. Your store becomes your profit channel.
• Don't move a commodity. If your only edge is price, your own store has nothing to add.

How to do it without wasting a year

On your own store, the number I watch is revenue per visitor: conversion rate times average order value. It tells you what every click is worth, which matters because you'll be paying for those clicks with ads. The worked example on our homepage: a 1% conversion rate at a $50 average order value is $0.50 per visitor, or $5,000 on 10,000 visitors. Move to 3% at $80 and it's $2.40 per visitor, or $24,000 on the same 10,000 visitors. That's 4.8 times the revenue from the same traffic.

That's why the product page is where the money is. The mistake Amazon sellers make on Shopify is copying the Amazon listing over word for word. An Amazon listing is written for a buyer who's already comparing five options on one screen. A Shopify page has to do the whole job: explain, prove, answer objections and build the bigger order.

And your Shopify buyer probably has an Amazon tab open. Plan for it. We wrote a full guide on writing a Shopify product page for Amazon shoppers that shows how to win that price check instead of pretending it isn't happening. For the side-by-side economics of both channels, read Shopify vs Amazon revenue per visitor.

If you want the whole story of why Amazon sellers are moving their winners to Shopify, it's on our page for Amazon sellers.

Your Q4 readiness checklist

Print this. Tick it off before October 14.

Start here: the full reading list

This guide is the hub for everything we've written for Amazon sellers. Here's where to go next, in the order I'd read them.

Fix the listing first

• Amazon product page conversion: fix the leak, the two-number view of your listing
• 50 Amazon listing audits and the flaw in 80% of them
• What an Amazon listing audit service should find
• Amazon bullet points: the conversion rate study

Then build the second shelf you own

• Shopify vs Amazon revenue per visitor
• How to write a Shopify product page for Amazon shoppers
• How to write a product page for TikTok traffic
• How to write a product page for Black Friday traffic
• Why Amazon sellers are moving their winners to Shopify

Free tools for this quarter

• Q4 restock calculator and storage fee forecaster
• FBA fee calculator and break-even ACoS calculator
• Coupon profitability calculator and PPC audit
• AI image policy checker and listing images grader

FAQ

Is Rufus still on Amazon in 2026? Not under that name. Since May 13, 2026, it's part of Alexa for Shopping, which answers questions in the search bar, writes AI overviews, compares products and shows up to a year of price history.

What are Amazon's 2026 holiday peak fulfillment fees? They run October 15, 2026 to January 14, 2027 and average $0.32 per unit on top of standard fees, across FBA, Remote Fulfillment, Multi-Channel Fulfillment and Buy with Prime. The 3.5% fuel and logistics surcharge still applies on top.

When is the Q4 2026 FBA inventory deadline for Black Friday and Cyber Monday? October 14 for AWD, October 21 for FBA with minimal splits, October 28 for FBA with Amazon-optimized splits. Those are arrival dates.

What is a good ACoS on Amazon in 2026? One below your profit margin before ad spend. A product keeping $12 on a $40 sale breaks even at 30%, and peak fees push that lower in Q4.

Can I use AI-generated images on Amazon? Yes, if they accurately show the product and follow the image rules. Retouching counts as normal editing, photorealistic AI-generated people must be disclosed on new uploads, and the main image stays on pure white.

Should Amazon FBA sellers start their own Shopify store? If you have a proven product that bundles, yes. Three client products went from $3.83, $7.73 and $17.58 net profit per sale on Amazon to $48.27, $78.87 and $134.28 on their own store, after paying for ads and inventory.

What to do next

Open Seller Central today and pull the fee preview for your top 3 products. Add the $0.32 average peak fee and the 3.5% surcharge, recalculate break-even advertising cost of sales for each one, and write the three numbers on a sticky note next to your screen. Every deal, bid and restock decision between now and January gets checked against that note.

---

Book Your Profit Audit

If one of your Amazon products already sells, it's worth knowing what it would earn on a shelf you own. Get your free profit audit and we'll run that math with you, then show you how to rebuild a high-converting product sales page in less than 15 minutes.

Book Your Profit Audit →

Or go here to check it out → revenueflows.ai

P.S. Amazon raised the rent twice this year and hired an AI to pick who gets the best shelf. Keep renting if it pays. Just make sure you own one store where nobody can raise it.