Online Arbitrage on Amazon: The Complete 2026 Guide

Look, real talk. There are about 400 blog posts on the first page of Google for "online arbitrage on Amazon" and most of them are written by people who have never shipped a single FBA box. I've been selling on Amazon since 2018. I do over $100K a month. I run a coaching program with about 70 active students who are scaling past $10K a month with this exact model. So when I write a "complete guide," it's the actual playbook, not the recycled version you've already read four times.

Here is what we're going to cover. The real definition. The math. The sourcing stack I use every day. Keepa. The ROI fight. Ungating. The buy box. Prep. Replens. And the failure modes that kill most beginners in their first 90 days. By the end of this you will have a working mental model of the entire business, end to end, with no fluff.

If you want to skip the reading and watch me run the entire system live on a real ASIN, I do a free training every Thursday at 8 PM EST. Reserve your seat here. You'll see sourcing, analysis, and the actual buy decision, all in 60 minutes.

1. What Online Arbitrage Actually Is

Online arbitrage on Amazon means buying products at a lower price from one online retailer, then reselling them on Amazon at a higher price, and pocketing the difference after Amazon fees. That's it. You don't create the product. You don't own the brand. You don't negotiate with a Chinese supplier. You exploit a price gap that already exists between two marketplaces.

The way most people frame it gets the model wrong. They think online arbitrage is about "finding deals." It's not. It's about finding repeatable price gaps. A deal is something you buy once. A repeatable price gap is something you buy every week for 18 months. The first one pays for groceries. The second one builds a business.

Here's a concrete example. Say Target has a private-label-brand vitamin on clearance at $12.99 with a 20 percent Circle offer, so you pay $10.39. The same exact ASIN sells on Amazon for $24.99. Amazon takes a 15 percent referral fee plus an FBA fulfillment fee of about $4.20. Your net after fees is roughly $16.55. Subtract your $10.39 cost and you're at $6.16 of net profit per unit. That's 59 percent ROI. If you can buy 30 of them this week and that price gap holds for 6 months, you've got a $1,100/month income stream from one ASIN.

Online arbitrage isn't dropshipping. The product moves through your hands (or your prep center's hands) before Amazon ever sees it. It isn't retail arbitrage either, even though they're cousins. I've written a full comparison on online arbitrage vs retail arbitrage if you want the breakdown. The shortest version: OA scales because you do it from your laptop. RA hits a ceiling fast because you can only walk through so many stores.

2. Why Amazon, Why Now: The 2026 Numbers

You'll see a lot of people on YouTube say "Amazon is saturated" in 2026. They're wrong, and the data is the loudest argument against them. Amazon is still about 38 percent of US ecommerce. There are roughly 600 million ASINs in the US catalog. The number of products with a real, repeatable, FBA-friendly price gap at any given moment is probably in the seven figures. You don't need a million. You need 30 to 60 active replenishable ASINs to hit $10K a month at typical ROIs.

What changed in 2026? Three things. First, Amazon raised fees twice in the last 18 months, which compressed margins for new sellers and weeded out the laziest competition. Second, the IPI score thresholds got stricter, which means slow-moving sellers are getting their inventory capped. Both of these are good for serious operators because they shrink the pool of competitors. Third, more retailers (Kohl's, JCPenney, Belk, Boscov's) are using aggressive clearance + coupon stacking as a way to compete with Amazon directly, which creates the exact price gaps OA sellers exploit.

The result: the model is harder for tourists and better for operators. If you treat it like a real business with real systems, you win. If you treat it like a side gig and skim YouTube tutorials, you'll lose. That's been true forever, but in 2026 the gap between the two is wider.

3. The Day-One Math: How Much Money You Actually Need

You will read everywhere that you can start online arbitrage with $100. Technically true. Practically, garbage advice. Here is what you actually need before you place a single order:

• Amazon Professional seller plan: $39.99/month. Don't bother with Individual. You can't win the buy box on Individual.
• Keepa subscription: 19 euros/month for the full data set. This is non-negotiable. You can't analyze ASINs without the pink buy-box line.
• SellerAmp SAS or equivalent scanner: about $20/month after the trial.
• Starting inventory budget: $500 to $1,000 at minimum. More is better.
• Prep budget: roughly $1.30 to $1.80 per unit if you use a prep center.

So your "real" day-one out the door is closer to $700 to $1,200 once you account for tools and the first round of inventory. I've written a much deeper breakdown on how much money it actually costs to start Amazon FBA with specific category breakdowns, but those are the numbers. If you have less than $500 you can still start, and I wrote a full guide on starting online arbitrage with $500 for that exact case. Just understand the constraint: under $500 you can't stay diversified enough across ASINs to survive normal variance.

The willpower-fix trap most beginners fall into is thinking they need more capital. They don't. They need a system that turns their existing capital over faster. A $500 buy at 35 percent ROI that sells in 30 days does $175 of profit per cycle. Run that 12 times in a year and your $500 produced $2,100 of profit. The lever isn't your starting cash. The lever is your cycle time.

4. The Sourcing Stack: Where Real OA Sellers Find Products

Most beginners ask "where do you source?" expecting one magic answer. There isn't one. There's a stack. The stack rotates through five primary methods, and each one fills a different gap in your portfolio.

Method 1: Storefront stalking

You find a seller on Amazon who has a clean storefront of profitable ASINs, you load their ASIN list into a reverse-sourcing tool, and you trace which retailers sell their products cheaper. This is the fastest way to bootstrap a beginner portfolio because you're piggybacking on a more experienced seller's research. The downside: everyone else is doing the same thing, so margins compress fast.

Method 2: Reverse sourcing with Keepa

You use Keepa's product finder to filter for specific criteria (sales rank under 100K, buy box price above $20, FBA seller count under 8) and you generate a list of candidate ASINs. Then you go check each retailer's site for that product. Slower than storefront stalking but produces leads no one else has.

Method 3: Retailer crawling

You go to a retailer site directly (Kohl's, JCPenney, Target, Boscov's, Macy's) and you click through the clearance + coupon-eligible sections, paste each ASIN into your scanner, and check the math. This is the slowest method per lead but it surfaces stuff no tool will ever find.

Method 4: Target lists / lead lists

You subscribe to a curated daily list (paid or your own team-built list) of pre-vetted leads. Less leverage on margin but huge time savings.

"My VAs are working on it every single day, all day, adding sales that we can actually source for online arbitrage, right? It's because if you want to make money with online arbitrage you need to be looking at deals every single day." — Chris, Sourcing Profitable Online Arbitrage Deals LIVE | Amazon FBA Product Research (Jun 2024)

Method 5: Rabbit-trail / frequently-bought-together

You find one profitable ASIN, you look at what Amazon shows as "frequently bought together" or "customers who bought this also bought," and you check those for profitability. The logic: if other OA sellers are buying the original lead, they probably haven't gotten to its neighbors yet. This is one of my favorite leverage moves and I run it on every winner I find.

The stack matters because no single method will keep you fed. Storefront stalking burns out as competitors find the same storefronts. Retailer crawling is slow. Lead lists get saturated. You rotate. For a deeper walkthrough of the actual click-by-click process, see my full guide on how to find profitable online arbitrage products.

5. Keepa: The One Tool You Don't Skip

If you take one tool away from me I quit the business. Keepa is the price-history and sales-rank database for Amazon. It tells you what an ASIN really sold for over the last 6, 12, and 24 months. Without it you're guessing.

The graph has four lines that matter for OA: the Amazon line (orange), the new-third-party-FBA line (blue triangles), the new-third-party-FBM line (blue squares), and the pink buy-box line which is only visible on the paid version. Most beginners look at the orange line and miss the entire story. The buy-box line is what your competition is actually winning sales at. That's the line you need to match or beat to win the box yourself.

"There's a gap in between the two because FBA is always higher than FBM, right? Of course. Of course. Trying to get the margins." — Chris, The Last Keepa Tutorial You Will Ever Need (Apr 2025)

Three Keepa habits that separate winning sellers from losing ones:

1. Look at 90-day buy-box average, not current price. The current price is a snapshot. The 90-day average is the price you'll actually sell at across many sale cycles.
2. Check the sales rank chart, not just the rank number. A rank of 50,000 means nothing if it's been flat. A rank that bounces up and down (the "drops") tells you actual sales velocity. More drops = more sales per month.
3. Watch the offer count chart. If FBA seller count is climbing fast, the buy box is about to get more competitive and the price will probably drop. If FBA seller count is shrinking, you've found a window.

If you want the full tutorial with screenshots, I wrote one: how to read Keepa graphs. Bookmark it and reread it every time you sit down to source for the first month.

6. ROI vs Margin: The Question That Eats Beginners Alive

This is the single most-asked question I get from new students. ROI vs margin, which one matters? Most people use them interchangeably. They're not the same thing.

"Return on investment, profit margins, these two are often confused amongst beginners. They use one in case of the other but they are two different KPIs. They tell you a different story. They are both important but on your day-to-day activities you are going to look at one more than the other." — Chris, ROI vs Profit Margins: What's More Important for Amazon FBA (Jun 2023)

The math:

• ROI = net profit ÷ cost of goods. Tells you how hard your money is working.
• Margin = net profit ÷ sale price. Tells you what slice of each sale you keep.

Same example, different framings. You buy something for $10, sell on Amazon at $25, and net $5 after fees. ROI is 50 percent ($5/$10). Margin is 20 percent ($5/$25). Both true. Both useful for different reasons.

For online arbitrage product analysis at the buy decision, ROI is the decisive number. We want at least 30 percent net ROI on every buy, ideally 40 to 50 percent on the meat of the portfolio. Margin matters for full-portfolio reporting and for understanding your cash flow at the business level, but at the unit level you live and die by ROI.

Why? Because OA is a cash-cycle game. Your cash goes out, sits in inventory for 14 to 60 days, then comes back as a payout 14 days after the sale. If you ran a 30 percent margin business with 90-day cycle times you'd still make less money than a 30 percent ROI business with 21-day cycle times, because the second one gets to spin its capital 4x faster. Margin doesn't fund your next buy. ROI on a fast-turning unit does.

7. Getting Ungated in Profitable Categories

About 70 percent of the highest-ROI ASINs I source are in categories or brands that require ungating. Beauty, toys, grocery, baby, top-tier brands across all categories. If you don't ungate, you're sourcing from the 30 percent of the catalog where everyone else is sourcing too. That's why "saturation" is mostly a self-imposed problem.

The fastest path to ungating in 2026:

1. Source a lead in a gated category as if you weren't gated. SellerAmp will tell you it's gated.
2. Buy a small wholesale lot (usually 10 units minimum, $200 to $400 budget) from a legitimate distributor with a real invoice. The invoice has to match Amazon's strict format: your business name, real address, supplier with verifiable contact info, 10 units minimum, dated within the last 180 days.
3. Apply for ungating directly inside Seller Central using that invoice. Amazon usually responds within 24 to 72 hours.

Some categories like Beauty, Grocery, and Toys unlock with a single approved invoice. Some brands (especially the high-protection ones) require Brand Registry letters which are functionally impossible for resellers. Skip those and move on. There are thousands of unlocked brands inside each gated category and you don't need the protected ones.

The deeper walkthrough with screenshots and Amazon's exact invoice requirements is over here: how to get ungated on Amazon: the master guide. It's the second-most-bookmarked post on this blog for a reason.

8. The Buy Box and Why You Lose Sales Without It

The buy box is the little box on the right side of an Amazon product page with "Add to Cart" and "Buy Now." Whoever has the buy box gets about 99 percent of the sales on that ASIN. If you don't have it, you sit on inventory.

"Winning the buy box is super important because all the sales happen in the buy box. 99 percent of the time, not all of them, but all of them, let's just call it all the sales are happening at the buy box level. You do not need to be the cheapest price. You do not need to have the most reviews. All this plays a role in the algorithm because, yes, it is an algorithm that basically dictates who wins the buy box." — Chris, Watch This to Master The Amazon Buy Box Algorithm (May 2023)

The buy-box algorithm weighs these factors (Amazon will never publish exact weights, this is from years of watching the behavior):

• Price (relative to competitors, not absolute)
• Fulfillment method (FBA almost always beats FBM at the same price)
• Seller performance metrics (Order Defect Rate, Late Shipment Rate, Cancellation Rate)
• Account health and tenure
• Stock level (out of stock = no buy box)
• Shipping speed (Prime two-day eligibility)

The non-obvious trick: you rotate the buy box with competitors when prices are within a few cents of each other. Amazon splits time. Most beginners don't realize this and panic-undercut, which destroys the price for everyone. Don't undercut. Match. The split-time mechanic does the rest.

9. Prep, Ship, Repeat: From Order to Amazon Warehouse

You bought the inventory. Now what? You have two paths.

Path A: Self-prep at home

You ship the units to your house, you label each one with a poly bag if needed, you apply Amazon's FNSKU sticker, you box them, you generate a shipping plan in Seller Central, you print the FedEx or UPS label, and you drop the boxes at the carrier. Time per shipment: roughly 30 minutes to 2 hours depending on unit count. Cost per unit: about $0.30 in supplies, but the real cost is your time.

Path B: Prep center

You ship the inventory directly from the retailer to a prep center. They receive it, inspect it, label it, polybag it if needed, ship to Amazon. You never touch it. Cost: roughly $1.30 to $1.80 per unit. Time per shipment for you: 5 minutes (you forward tracking numbers).

For sellers under $5K/month I tell them to self-prep at home for the first 90 days. You need to understand the process before you outsource it. Once you cross $5K/month consistently, switch to a prep center immediately. The hour you save per shipment is worth more than $1.30 per unit. I wrote a deeper comparison: Amazon prep center vs DIY prep at home.

10. The Replenishable Lead: How One ASIN Becomes $50K a Year

This is the most underrated concept in the entire model and the reason real OA sellers crush hobbyists.

A "lead" is one buy of one ASIN. Most beginners get excited about finding leads. The actual gold is finding replens, which is short for replenishables: an ASIN where the price gap holds for months and you can keep buying it over and over again.

"Right now I've dedicated a lot of my sourcing, so I'm not sourcing as much as I used to for online arbitrage. The reason why I haven't been doing a lot of online arbitrage sourcing is because I'm dedicating a lot of the online arbitrage part of the business to my team and my prep center doing my prepping. So that's a big part of my business." — Chris, Live Online Arbitrage Sourcing: $120 of profit in 30 min (Replen) (Apr 2024)

One ASIN at $8 profit per unit, selling 20 units a month, replenishable for 12 months = $1,920 per year from one ASIN. Find 30 of those and you're at $57,600/year of profit from a portfolio of 30 ASINs you sourced once and rebuy on schedule.

That's the actual job of online arbitrage. Not finding 500 one-off leads. Finding 30 to 60 replenishables and feeding them every week. My personal portfolio is mostly replens at this point. Most of my new "sourcing" time is actually about defending and expanding my existing replen list, not finding fresh leads. The day-in-the-life of a 6-figure online arbitrage seller is mostly replenishment, not discovery.

11. Why Most Online Arbitrage Sellers Fail

I've watched probably 800 people start online arbitrage in the last three years between students, viewers, and people in the Discord. The failure modes are almost always one of these seven:

1. They never source enough. They look at 20 ASINs in a day and decide "nothing is profitable." Sourcing is a numbers game. You scan 150 to 200 ASINs to find 5 buys.
2. They don't trust Keepa. They buy on hope. The Keepa data was screaming "this drops in 30 days" and they bought anyway.
3. They chase ROI over replenishability. A 70 percent ROI one-time buy is worse than a 35 percent ROI replenishable.
4. They never ungate. They stay in the 30 percent of the catalog where everyone else is.
5. They scale too slow. They keep their portfolio under 10 ASINs forever and wonder why revenue is flat.
6. They scale too fast without systems. They throw $20K into inventory before they've validated their sourcing or their prep stack, and they end up with stranded inventory and IPI score issues.
7. They confuse activity with progress. 6 hours on Reddit reading about OA isn't sourcing. 1 hour scanning ASINs is.

I went deep on each of these in why online arbitrage sellers fail: 7 real reasons, including the exact fix for each one. If you've already been doing this for a few months and you're not where you want to be, that post is the one to read.

12. Your 30-Day Action Plan

Stop reading. Start moving. Here is the actual 30-day sequence I'd run if I were starting today with $1,000.

Week 1: Setup

• Open the Amazon Professional seller account ($39.99/mo).
• Subscribe to Keepa (paid version, 19 euros/mo).
• Subscribe to SellerAmp SAS (~$20/mo after trial).
• Set up a dedicated business bank account and credit card.
• Pick a prep center or set up a home prep workflow (boxes, scale, printer).

Week 2: Source

• Run storefront stalking on 3 to 5 OA sellers' storefronts. Generate a list of 200 candidate ASINs.
• Analyze each one in SellerAmp + Keepa. Keep only ones at 30 percent+ net ROI, 90-day buy-box average above $15, sales rank under 100K, FBA seller count under 10.
• Your goal: 5 to 8 verified buys.

Week 3: Buy and Ship

• Place the orders. Use rebate-portal cashback (TopCashback, Rakuten) to add 2 to 8 percent on top of your ROI.
• Ship to your prep center or to your house. Label, polybag, box, ship to Amazon.
• Total spend: probably $400 to $600 in inventory.

Week 4: Reload

• While Week 3's units travel to Amazon and get checked in, start sourcing your Round 2.
• Apply for ungating on at least one category. Start with Beauty (easiest).
• Set up your repricer. Read this on how to use a repricer first.

By Day 30 you'll have your first 5 to 8 ASINs live, sales coming in, your first cash payout in 14 days from your first sale, and a list of Round 2 leads ready to buy. That's a real start. Not a fake start. Not a tutorial start. A real OA business.

Where to go next

Online arbitrage is one of three Amazon paths. If your goal is closer to a structured FBA business, the next pillar to read is how to start an Amazon FBA business in 2026. If you're already at $5K to $10K a month and trying to break past the $10K ceiling, read how to scale your Amazon FBA business past $10K a month.

And if you want to skip the slow road, the highest-leverage move I can give you is to spend 60 minutes on the free training I run every Thursday at 8 PM EST. I show the entire $200K/month system live, including a real sourcing session, a real Keepa analysis, and the buy decision in real time.