Amazon, Walmart or Your Own Store: What Each One Really Costs

Every product business gets to this question eventually. You've got something people want to buy, and two ways to sell it. You can put it in front of the enormous crowd already shopping on Amazon and Walmart, or you can send people to a store you own. Most advice picks a side and argues it hard.

The honest answer is that they do different jobs, and the mistake is not choosing wrong. It's choosing without knowing what each one actually costs. Marketplaces rent you customers by the sale. Your own store makes you go find them, then lets you keep them. Both of those are real bills, just paid in different currencies.

Here's what each channel costs in 2026, what you give up, and how to decide without guessing.

The short version

• Amazon referral fees run 8 to 15 percent for most categories and did not change for 2026, but total costs including fulfillment commonly land in the 30 to 45 percent range of the selling price.
• Walmart charges no monthly fee and referral fees from 6 to 15 percent, which makes it the cheaper marketplace for many categories.
• Your own store has lower fees and higher work. You pay for traffic with money or time instead of paying commission.
• The bigger difference is not the fee. It's that a marketplace keeps the customer relationship and you cannot market to those buyers again.
• Most product businesses should run both, with the marketplace as discovery and the store as the place repeat customers land.

What you are actually renting on a marketplace

When you list on Amazon or Walmart, you are not renting shelf space. You're renting demand. Somebody is already there with a credit card saved, already trusting the checkout, already expecting fast delivery. You didn't earn any of that and you don't have to.

That's worth real money, and the commission is the price. A 2024 Jungle Scout survey of 1,000 US online shoppers found 56 percent started product searches on Amazon compared with 42 percent on search engines and 29 percent on Walmart.com. That figure moves around by survey and has softened since its peak, but the direction is consistent. When somebody already knows they want to buy, a lot of them skip Google entirely.

The catch is what you're renting rather than owning. You get the sale. You do not get the customer. Amazon and Walmart both sit between you and the buyer permanently, and that arrangement does not end when the order ships.

What an Amazon sale actually costs in 2026

Amazon announced its 2026 changes ahead of time, and the news was mild. Fulfillment by Amazon fees rose an average of eight cents per unit, effective January 15, 2026, which Amazon describes as less than half a percent of an average item's price. Referral fees did not change at all, and no new fee types were introduced. That follows a year with no increases.

So the structure most sellers are working with looks like this.

• Referral fee. A percentage of the sale price, ranging from 8 to 45 percent by category, with most categories at 15 percent. This is the unavoidable one.
• Professional selling account. $39.99 per month, worth it above about forty units a month.
• Fulfillment. If you use Fulfillment by Amazon, roughly $3.22 and up per unit depending on size and weight.
• Storage. Charged per cubic foot per month, and it climbs sharply in the fourth quarter.
• The ones that surprise people. Inbound placement, returns processing, low inventory level fees. Published estimates put these at another one to three dollars per unit.

Add it up and the commonly cited range for total Amazon costs is 30 to 45 percent of the selling price. Treat that as a planning range rather than gospel, because it swings hard on product size, weight and category. The reliable move is to run your actual product through Amazon's own Revenue Calculator before you list anything.

Is that expensive? Compared to a percentage, yes. Compared to what it would cost you to build the same demand from nothing, that's a much harder question, and it's the right one to ask.

What a Walmart sale costs, and why it is often less

Walmart Marketplace is the cheaper of the two on paper, and it isn't close in some categories.

There's no monthly subscription, no setup fee and no listing fee. You pay a referral fee when something sells, and those run from 6 to 15 percent. Personal computers sit at 6 percent. Appliances, electronics and video game consoles at 8 percent. Automotive, power tools and industrial at 12 percent. Books, home and kitchen, and toys at 15 percent. Several categories including apparel, baby, beauty and health, jewelry and watches are tiered by item price.

Walmart's fulfillment service charges from $3.45 for items up to a pound, with storage at $0.75 per cubic foot in the January to September window. There's also a New Seller Savings program running February 2, 2026 through January 31, 2027, discounting referral fees by 20 percent on your first $50,000 in sales and more at higher volumes.

The tradeoff is traffic. Walmart's marketplace is real and growing, but it is not Amazon-sized, and the seller tools are less mature. Cheaper fees on fewer customers is a genuine tradeoff rather than a free win.

A marketplace rents you customers by the sale. Your own store makes you go find them, and then lets you keep them.

What your own store really costs

This is where most comparisons get dishonest, because they list the platform fee and stop.

The direct costs are genuinely low. A hosted store platform runs somewhere in the range of a few hundred dollars a year. Payment processing takes roughly 2.9 percent plus about thirty cents per transaction, which is the standard published rate for most card processors. Add a domain and whatever apps you actually need. For most small sellers the hard costs land well under ten percent of revenue.

Then comes the part nobody quotes. Nobody knows your store exists.

• Traffic. You pay for it with advertising money, with SEO work over months, with email, with social, or with your own time. This is the real cost and it is not optional.
• Trust. A shopper who has never heard of you needs reasons to hand over a card. Reviews, clear policies, a real address and a site that doesn't look improvised.
• Operations. Shipping, returns, customer service and the software that ties it together. On a marketplace, some of that comes with the commission.

So the fee comparison is misleading on its own. The right comparison is your total cost to acquire and serve one customer on each channel. If you're spending twenty-five percent of revenue on ads to make sales on your own site, you have not beaten a fifteen percent referral fee. You've just moved the money and taken on more work.

Before you shift budget between channels, know what a customer actually costs you to acquire. Our free calculator does the paid-ads side of that math in about two minutes.

Run the ROI numbers

The thing a marketplace takes that is not money

If the whole question were percentages, most small sellers would stay on marketplaces and be content. It isn't, and this is the section worth reading twice.

On a marketplace, the buyer is the platform's customer, not yours. You typically don't get their email address in a usable form. You can't add them to a list, send them a launch announcement, offer them a reorder discount, or build a lookalike audience from them. Ask any seller who has tried to move a product line off Amazon after five good years how many of those buyers followed. The number is disappointing, and it isn't their fault.

That matters because the second sale is where the profit lives. Industry benchmarks on repeat purchase rates vary widely by category and should be read as rough guides rather than precise figures, but the direction is not in dispute: a customer who buys twice is dramatically more valuable than one who buys once, and you cannot pursue a second sale from someone you cannot contact.

There's also the risk nobody plans for. A suspended listing, a category fee change, a policy shift, or a competitor hijacking your product page can take your revenue to zero on a Tuesday with no warning. If a marketplace is your only channel, the platform holds the business. Your own store is the part nobody can switch off.

How to decide, honestly

Skip the ideology. Answer these.

1. Do people already search for what you sell? If your product is a known category with real search volume, a marketplace puts you in front of that demand tomorrow. If you sell something people don't know exists, a marketplace has nobody looking for it and your own store plus targeted advertising is the better path.
2. What's your margin after commission? Run one real product through the numbers. If a 15 percent referral fee and fulfillment leave you nothing, that product does not belong on a marketplace at that price, and no amount of optimization changes the arithmetic.
3. Does anyone buy from you twice? If your product is consumable, replaceable or part of a range, the customer relationship is worth more than the commission you saved, and your own store matters a great deal. If it's a genuine one-time purchase, that argument weakens.
4. Is your brand the reason people buy? On a marketplace you sit next to eleven competitors on a page designed for comparison. If your advantage is story, service or expertise, that page is the worst possible venue for it.
5. What can you actually operate? Two channels is more work than one. A well-run marketplace listing beats a neglected store, and the reverse is just as true.

Across North Texas the pattern I see most often is a business doing one of these properly and the other one badly, usually because the second channel got set up during an enthusiastic week and then left alone. Half-running two channels is worse than running one well.

Running both without doubling the work

For most product businesses the answer is both, in specific roles. The marketplace is where new customers find you. Your own store is where the ones who liked you come back.

Making that work takes a few decisions up front.

• Keep pricing consistent. Undercutting your own marketplace listing looks like a bargain and reads as untrustworthy. Compete on bundles, service and range instead of on the same item at two prices.
• Write your product content once, properly. Titles, bullets, descriptions and product images carry over between channels with editing rather than rewriting. Getting the source content right is where the effort pays twice.
• Let one system own your stock. Selling something twice because two channels disagreed about inventory is the fastest way to a suspension and a bad review at the same time.
• Put something in the box. A packing insert is the one legitimate bridge from a marketplace order to your own store. A card offering a warranty registration, a care guide or a reorder discount converts a rented customer into one you can reach.
• Watch profit by channel, not revenue. Revenue on a marketplace flatters itself. Track what you keep.

That fourth point is the whole strategy in miniature. Marketplaces are excellent at introducing you to people. They're just not willing to hand those people over, so you have to earn the handoff in the packaging.

A ninety-day plan

1. Pick one product and do the real math. Selling price, cost of goods, referral fee, fulfillment, shipping, returns. What you keep per unit on each channel. Everything else is opinion until this exists.
2. Fix whichever channel is already half-built. Usually the store. Clear product photos, honest shipping and returns policies, a working search, and a checkout you have personally tested on a phone.
3. Get your product content right once. Rewrite titles and bullets to answer what a buyer actually needs to know, then adapt to each marketplace's format.
4. Add structured data to your own store. It's how search engines and AI systems read your prices, stock and reviews, and it's free.
5. Start the packing insert. Cheapest customer acquisition available to a product business, and most sellers never print one.
6. Build the email list from day one. Every store order, every insert signup, every trade show card. This is the asset no platform can take.
7. Review profit by channel quarterly. Fees change, categories get reclassified, and a product that worked at last year's rates might not work at this year's.

If your own store is the weak half, our website design and development work usually starts there, and the jobs your website is turning away covers the friction points that quietly cost sales.

---

Frequently asked questions

Is it cheaper to sell on Walmart than Amazon?

Usually yes, on fees. Walmart charges no monthly subscription and referral fees of 6 to 15 percent, with personal computers at 6 percent and electronics at 8 percent, where Amazon charges 15 percent in many of the same categories plus $39.99 a month. The tradeoff is traffic volume. Cheaper fees only help if enough shoppers see your listing.

What percentage does Amazon actually take?

The referral fee is 8 to 45 percent by category, with most categories at 15 percent, and it did not change for 2026. Once you add fulfillment, storage and the smaller charges, published estimates commonly put total costs at 30 to 45 percent of the selling price. That varies a lot by product size and weight, so run your specific item through Amazon's Revenue Calculator.

Should I stop selling on Amazon and just use my own website?

Rarely, and not abruptly. Amazon reaches buyers you cannot easily reach on your own, and cutting it off usually means losing that volume before your own traffic is ready to replace it. The better sequence is to keep the marketplace running while you build your store, your email list and your repeat customers, then shift the balance gradually.

Can I get customer contact details from marketplace orders?

Not in a form you can market to. Both platforms deliberately sit between you and the buyer, and using order data for outside marketing violates their terms. The legitimate route is a packing insert that gives the customer a reason to come to you directly, such as warranty registration, a care guide, or a discount on their next order.

How much does it cost to run my own online store?

The hard costs are modest. Expect a few hundred dollars a year for the platform, roughly 2.9 percent plus thirty cents per transaction for card processing, and a domain. The real cost is traffic, paid for in advertising, SEO work or your own hours. Budget for that honestly or the store will simply sit there.

Not sure which channel is actually making you money? We'll look at your real numbers by channel and tell you where the margin is going, at no charge.

Get a channel review

We help product businesses in Texas run marketplaces and their own stores as one system rather than two competing projects. Our eCommerce and marketplace services page explains how that works in practice.

Sources

1. Amazon Selling Partners, 2026 Updates to US Referral and Fulfillment by Amazon Fees, effective 2026-01-15, retrieved 2026-09-08. sellingpartners.aboutamazon.com
2. Feedvisor, Amazon Referral Fees by Category, 2026 rates, retrieved 2026-09-08. feedvisor.com
3. Feedvisor, Walmart Seller Fees: Referral, WFS and Storage Costs, 2026, retrieved 2026-09-08. feedvisor.com
4. Webgility, Marketplace Fees 2026: Amazon, eBay, Etsy and Walmart Seller Costs Compared, retrieved 2026-09-08. webgility.com
5. Jungle Scout via Search Engine Land, 50% of product searches start on Amazon, survey of 1,000 US shoppers, 2024, retrieved 2026-09-08. searchengineland.com