Leave Amazon for your own store without losing the sales you already have

The commission is not a tax on your listing; it is the price of a bundle, and sellers usually underrate what is in it. The platform brings buyers already in a purchasing mood, lends you a trust signal you have not earned yet, runs the payment rails and the fraud checks, and absorbs the dispute when something goes wrong. It also handles the parts of commerce nobody enjoys: chargebacks, refund arguments, and sales tax, which Amazon collects and remits as the marketplace facilitator in every US state that has one. When you leave, none of that disappears. It moves onto your desk. Whether the bundle has value is not in question; whether it is worth its percentage at your volume, and whether you can assemble a cheaper version yourself, is. Published rates are the place to start. Amazon's referral fee runs roughly 8-15% by category, with 15% the most common rate and a $0.30 minimum per item in most categories, and the Professional selling plan adds $39.99 a month. Payment processing is folded into the referral fee, which is the one way Amazon is cheaper than it looks.

• Demand: buyers already searching with intent, on a channel you did not build
• Trust: a return policy and a brand name shoppers already accept
• Payments: processing, fraud screening and US sales tax collection
• Disputes: chargebacks and refund arguments handled by someone else

The comparison sellers make is wrong. They compare the marketplace commission to nothing, as if traffic on their own domain arrives free. The real comparison is commission against blended customer acquisition cost. Use the fee schedules quoted above, not a forum estimate, and remember the referral fee comes before FBA fulfillment fees, which are charged per unit by size and weight. Add the Sponsored Products spend you already buy inside the platform, because most sellers are paying twice; Amazon sellers track it as TACoS, ad spend divided by total sales, and that number belongs in the comparison. Now price your side: ads, content, email tooling, and the build itself, which for a serious storefront starts in the low four figures — our pricing page is blunt about the range. If your own acquisition cost per order lands below the all-in marketplace take, the move pays. If it lands above, you are buying independence, not margin, and you should know which one you are buying.

The store is the easy half. The hard half is that nobody arrives on day one, and no amount of design fixes that. On Amazon, search was a feature bundled into the referral fee. On your own domain, search is a project: category and product pages that can actually rank, technical hygiene, and the patience for organic traffic to compound over quarters, the unit it works in, which is what SEO work really means. Until then you pay for demand with ads, content, social, partnerships or an email list you have been building. Sellers who leave without a demand plan do not discover a cheaper business; they discover a quieter one. Do the sum: if Amazon takes 15% and you would spend 20% of revenue buying the same volume through ads, leaving costs you money until the organic side compounds.

Switching cold is the most common self-inflicted wound. Keep the Amazon listings running at full volume and treat your own store as a second channel that has to earn its share. If your stock already sits in FBA, you do not need a second warehouse for it: Amazon's Multi-Channel Fulfillment ships orders placed on your own site from that same inventory, in unbranded packaging if you ask for it, at its own per-unit rate. Start with the products where independence matters most: your repeat-purchase items, where a returning customer is worth far more than a single commission, and your best sellers, where the fee is largest in absolute dollars: at a 15% referral rate an $80 order hands over $12 every time it repeats. Leave the long tail on Amazon, where its search still does useful work for products you would struggle to get traffic to. Give the parallel period real time — several quarters, not several weeks — and let the numbers, not the frustration with fees, decide how far the shift goes.

• Keep Amazon listings at full volume during the transition
• Move repeat-purchase and high-margin best sellers first
• Leave the long tail where Amazon search still finds it
• Judge the shift on channel numbers, not on annoyance with fees

The asset Amazon withholds is the customer. Buyer email addresses reach you as anonymized relay addresses, and Buyer-Seller Messaging is limited to messages needed to complete the order, so nothing you learn about an Amazon buyer can be used to market to them. Every order on your own domain, by contrast, should leave you with something durable: an email address with consent, an order history, a segment you can reach again without paying for the privilege. Start collecting with the first order, in an email tool such as Klaviyo or Mailchimp that reads WooCommerce order data, and connect it to measurement that shows which channel produces repeat buyers, which is what a working analytics setup is for. Post-purchase touchpoints are where independence gets built: packaging inserts, a real thank-you email, a reorder reminder timed to your consumption cycle. Read the platform rules before using any of them on marketplace orders. Amazon's seller policies are explicit that a card in the box pointing buyers to your own site is grounds for suspension, and no insert is worth that risk.

• Collect email with consent on every order from your own store
• Keep order history and segments in a system you control
• Use inserts and post-purchase email only where platform rules allow
• Measure repeat rate per channel, not just first-order volume

Independence arrives with a job description attached. Photography and copy are now yours to produce and yours to keep, which is an advantage but also a standing cost. Shipping stops being a label in a seller dashboard and becomes carrier accounts, rates, packaging and lost-parcel conversations. Returns need a written policy that a stranger will trust, and someone to process them. Sales tax becomes your obligation where Amazon used to collect on your behalf, which in the US means understanding your economic nexus: commonly $100,000 of sales into a state, sometimes a transaction count. Assuming it away is not an option, and the answer belongs to an accountant, not to a landing page. Support arrives in your inbox at the worst moments. None of this is a reason to stay, but it is the reason a store that looks cheap on paper is not free in practice.

On the software side the boundary is clean. WooCommerce runs the business: your catalog, stock, orders, coupons, tax rules, payment gateways and the plugins you already rely on stay exactly where they are, and your team keeps working in WordPress. What we rebuild is the customer-facing layer as a Next.js storefront, because that is the part responsible for speed, ranking and the buying experience. What you gain over Amazon is structural: no per-order commission, a checkout you control instead of one you rent, customer data that belongs to you, and the freedom to change price, bundle and presentation without asking permission from a rules page that can be rewritten at any time.

If every order you have ever taken came from Amazon search, leaving is a slow project, not a weekend. You are not moving a store; you are building a demand engine from zero while your existing one keeps the lights on. Do not start if margins are already thin enough that acquisition spend would erase them, if the catalog is a handful of low-repeat products nobody searches for by name, or if you have no capacity to answer support and process returns yourself. The cheaper first move is usually to test demand before you build: run a small paid campaign to a simple page, see what a click and a customer actually cost you, and only commission a storefront when that number makes sense. If the numbers say stay, stay, and revisit in a year.