DTC Ecommerce vs Amazon Marketplace: Where Each Channel Wins

For an established Amazon seller, DTC ecommerce is not cheaper than Amazon — it is a different bill. Amazon charges a referral fee that includes demand; a direct-to-consumer store removes that fee and hands you the cost of creating demand yourself. On a $29.99 single-unit product with typical fees, moving one order off Amazon frees up roughly $3.33 of margin, which is your entire budget for acquiring that customer. DTC wins when repeat purchases, order value, or an audience you already own push that number well past $3.33. It loses when every order has to be bought with a fresh ad click.

This guide prices both channels with figures checked on 2026-08-10, runs the same product through both, and gives four gates to decide with.

What DTC Ecommerce and Amazon Actually Cost in 2026

Start with the fixed and variable fees each channel publishes. These are the numbers that appear on invoices; the expensive lines come later.

Amazon marketplace (source: Amazon’s selling pricing page, data checked 2026-08-10):

Two things matter more than the numbers. Amazon deliberately does not publish a flat per-unit fulfillment figure, pointing sellers to its Revenue Calculator instead — so any comparison quoting a single FBA fee is quoting an example, not a rate. And the referral range is wide: most seller categories land between 8% and 17%, with 15% the most common figure. Category rates are in the Amazon referral fee guide; 2026 fulfillment and surcharge changes are tracked in the 2026 fee changes guide.

DTC store on Shopify (source: Shopify’s pricing page, USD-billed view, data checked 2026-08-10):

Shopify’s page also states plainly: “There are no setup fees on any of our plans.” One caution when you check this yourself: the pricing page localizes by country and currency and opens on the yearly-billing view by default, so one visitor’s screenshot will not match another’s. The figures above are the USD-billed columns as shown on 2026-08-10.

Then there is payment processing, which Amazon bundles into the referral fee and a DTC store pays separately. Stripe’s published US rate is “2.9% + 30¢ per successful transaction”, plus “1.5% for international cards” and “1% if currency conversion is required” (data checked 2026-08-10). Shopify Payments rates differ by country and plan; the third-party transaction fee above is what Shopify charges on top of your processor if you do not use Shopify Payments.

The Same $29.99 Product on Both Channels

Here is one unit of a standard-size $29.99 product, sold both ways. The referral fee uses 15%; the FBA fulfillment and storage lines are illustrative figures for a standard-size unit, not quoted rates — see the FBA fees and profit guide for how to derive yours.

Now the arithmetic. On Amazon the per-unit deductions total $10.40. On your own store, if you fulfill for roughly what FBA charges — a fair assumption for many sellers, and the reason Amazon offers Multi-Channel Fulfillment at all — your deductions are $5.90 of fulfillment plus $1.17 of processing, or $7.07.

The gap is $3.33 per order. That is the honest answer to “is DTC cheaper”: you keep $3.33 more per unit, and out of it you must pay to find the customer. Paid social and paid search rarely deliver a first-time buyer for $3.33 on a $30 product, so the single-unit, paid-traffic version of DTC loses on arithmetic, not on ideology.

That number is not fixed, though. Three levers move it:

• Average order value. Three units per order at $89.97 triples the freed-up referral fee. The break-even acquisition budget goes from about $3.33 to roughly $10, which is a different advertising market entirely.
• Repeat purchase. The $3.33 applies to every order, but the acquisition cost applies once. A consumable bought six times a year has a break-even budget near $20 per customer, not $3.33.
• Unpaid traffic. Brand search, email, content and an existing audience cost time rather than cost-per-click. The whole $3.33 stays margin.

If none of those three apply to your catalogue, the fee comparison has already answered your question.

Traffic Is the Line Item Nobody Puts in the Table

The referral fee is not a tax. It is the price of demand that already exists: a shopper who arrives on your listing was routed there by Amazon’s search, its ranking, and its checkout trust. Off Amazon, that shopper does not exist until you pay to create them.

This is also why Amazon will pay you to keep working both sides. Amazon’s Brand Referral Bonus offers, in Amazon’s own words, a bonus “averaging 10% of qualifying sales when you drive traffic to products in the Amazon store from search, social media, and other sources” (data checked 2026-08-10). Eligibility is narrow — Amazon states: “Sellers in Amazon’s US store with a Professional selling plan can enroll their brand in Amazon Brand Registry to be eligible for the Brand Referral Bonus.” The word doing the work is averaging: the rate is tiered by category and price, not flat, as broken down in the Brand Referral Bonus guide.

The practical read: if you are already buying off-Amazon traffic, Amazon will refund part of your referral fee for pointing it at Amazon, where conversion is highest. That makes external traffic a hedge you can test before you own a storefront, using someone else’s checkout to prove that your ads convert at all.

Data Ownership: What You Keep on Each Channel

On your own store you own the checkout, so the customer record — email address, shipping address, purchase history, product preferences — is yours by default, and so is the ability to email that person a refill reminder next quarter. That asset compounds; it is the single strongest argument for direct-to-consumer ecommerce and it does not appear anywhere in a fee table.

On Amazon the transaction is Amazon’s. You get order-level data through Seller Central reports and aggregate search behaviour through Brand Analytics, and buyer contact runs through Amazon’s own messaging system rather than a list you control. Rules on what you may send, and when, change periodically — check Amazon’s communication policy inside Seller Central rather than a third-party blog before building any post-purchase sequence.

The asymmetry to plan around: Amazon data tells you what sold, your own store data tells you who bought and whether they came back. Only the second kind supports a retention model, and a retention model is the only thing that makes the $3.33 math work.

Risk: Two Channels, Two Failure Modes

Neither channel is safe; they fail differently, and diversification is only meaningful if you know which failure you are insuring against.

Amazon’s failure mode is sudden and account-level. A suspension, a listing takedown or a compliance flag can stop all revenue in a day, with an appeals process you do not control — see the account suspension appeal guide. Listings can also be hijacked or altered by other sellers on the same ASIN (listing hijacking guide). Your store’s traffic is borrowed, and the lender can call it back.

A DTC store’s failure mode is slow and demand-side. Nothing gets taken away; nothing arrives either. A store with no traffic simply sells nothing while the subscription, the apps and the ad account keep billing. The failure is quiet enough that sellers keep funding it for months.

That difference dictates sequencing. A DTC store built while Amazon is healthy is insurance. A DTC store built in the week after a suspension is a startup with no runway and no traffic, launched under time pressure.

Four Gates Before You Open a DTC Store

Run these in order. A “no” at any gate is a reason to wait, not a reason to abandon the idea.

1. Do the economics survive paid acquisition? Calculate your own break-even figure the way the table above does: Amazon’s per-unit deductions minus your own fulfillment and processing costs. If the result is under $5 and you have no unpaid traffic source, stop here.
2. Is there a second purchase? Consumables, refills, and multi-unit bundles are what turn a one-time acquisition cost into a customer relationship. Single-purchase durables are structurally better on Amazon.
3. Do you have traffic you are not renting? Brand search volume, an email list, a content asset, or a real social audience. If the honest answer is “we will run ads,” you are entering an auction against brands with better margin per order.
4. Can you fulfill it? Pick, pack, ship, returns and customer service all move onto your books. Multi-Channel Fulfillment or a 3PL can absorb this, at a price you should compare against the FBA baseline before launch — the same comparison as FBA vs FBM, applied to a second channel.

Two practical notes on setup. Know your true Amazon unit economics before you use them as the baseline — the free AMZBase Chrome extension surfaces ASIN, BSR and lowest FBA offer on the listing page and calculates FBA fees to estimate potential profit, which is enough for a per-unit comparison without opening a spreadsheet. And your DTC store will need product media on day one: ASINCrate exports listing images and video from Amazon product pages in bulk with a CSV of metadata, with bulk ZIP export on its paid tier.

Frequently Asked Questions

Is DTC ecommerce cheaper than selling on Amazon?

Not on a per-unit basis for a single low-value order. In the worked example above, a $29.99 unit frees up about $3.33 by leaving Amazon, and that $3.33 has to cover customer acquisition. DTC becomes cheaper as order value, repeat purchase rate, or unpaid traffic increases.

How much can I afford to pay for a customer on my own store?

Your ceiling is the Amazon deductions you avoid minus the costs you add back — processing, fulfillment, storage, and the platform subscription spread over your monthly orders. Multiply that per-order figure by the number of orders you expect from one customer over a year, and you have a realistic acquisition budget rather than a per-order one.

Can I sell on Amazon and run a DTC store at the same time?

Yes, and most established sellers do. The two channels answer different questions: Amazon supplies demand and cash flow, the direct channel supplies customer data and a fallback. Running both also lets you use the Brand Referral Bonus to recover part of your referral fee on off-Amazon traffic while the store is still small.

What does a DTC store cost beyond the platform subscription?

Payment processing on every order, fulfillment and storage you now pay for directly, plus apps, a domain, and any advertising. The subscription is usually the smallest line item on the list; acquisition is almost always the largest.

Does an Amazon seller need a DTC store to build a brand?

No — Brand Registry, A+ content and Brand Analytics build brand assets inside Amazon. What a DTC store adds is a customer list you keep and a channel that survives an account issue. Treat it as insurance plus a retention channel, not as a cheaper way to sell the same units.

The Bottom Line

The DTC-versus-Amazon decision is not about which channel is fairer. Amazon sells you demand at a published rate and keeps the customer; a DTC store sells you the customer and makes you buy the demand. Price both on the same unit, as the table above does, and the choice usually resolves itself: single-purchase, low-value products stay on Amazon; repeat-purchase products, high order values, and brands with traffic they do not rent are the ones where direct-to-consumer ecommerce pays for itself. If neither applies yet, the useful move is not a storefront — it is testing off-Amazon traffic against your Amazon listings first, where the Brand Referral Bonus refunds part of the experiment.