You sell on Amazon. When does your own store pay off

The question about an own store usually comes back at one of two moments: after the first genuinely good quarter on Amazon, or after the first scare about the account. In both cases, “every brand should have its own store” is worth exactly as much as “do not touch what works”. The decision can be settled with numbers you already have in Seller Central.

Sources checked: 27 October 2026. Amazon rules and rates change independently of us, so check the current terms for your account and category before deciding.

What an own store will not fix

Amazon sells you access to demand that already exists. The customer arrives intending to buy, and you pay a referral fee on every order for that. An own store gives you a channel, but you have to bring the demand yourself: through ads, content, email, presence in search. That is the entire cost which, on Amazon, sits hidden inside the commission.

The second thing worth saying plainly: a store does not protect you from a suspension. When an account stops selling overnight, Amazon customers do not move to your store, because most of them do not know it exists. A store gives you a second leg to stand on, but only if you built it earlier and it already has traffic of its own. What to do when the account actually goes down is covered in our piece on Amazon account suspension.

If the idea still makes sense after those two paragraphs, that is a good sign. Below are the three situations where the arithmetic really closes.

Three situations where an own store starts to earn

1. The customer comes back for the same product

This is the strongest argument and the only one that survives a calculator. Cosmetics, household chemicals, pet food, supplements, consumables, spare parts, anything that gets used up. You pay the marketplace commission on every order, including the tenth order from the same customer. You pay the cost of acquiring a customer to your own store once.

You can check this without any tool: in your order report, count how many buyers appear more than once in a year and what they buy the second time. If repeats are rare, the rest of this article is trivia.

2. An offer Amazon cannot handle

Some things simply do not fit the marketplace model: custom bundles assembled per customer, subscriptions, personalisation (engraving, made-to-order size or colour), products with category restrictions, individual prices and terms for regular buyers. Amazon has its own channel for companies, covered in our piece on Amazon Business, but a wholesale price list with tiers, buyer limits and order approvals is a separate platform.

This reason gets confused with the first one. The difference: here the store does not replace Amazon, it handles sales that were never on Amazon at all.

3. A brand people search for by name

If buyers type your brand name, the demand is yours rather than rented. A store then captures traffic you currently pay commission on. Three places to check it: brand queries in Google Search Console, the share of branded terms in your Amazon Ads search term report, and whether anyone already pairs your brand name with the word “store”.

Mind the order: brand recognition is almost always built on Amazon and only then monetised in an own store. The reverse rarely works.

Run the break-even on one product, not on the whole company

The arithmetic that matters fits into three numbers: what is left from an order on Amazon, what is left from the same order in your own store, and what it costs to bring a customer to that store.

The Amazon referral fee depends on the category, starts at 2.5% and is charged on the full amount paid by the buyer, including shipping (sell.amazon.pl). FBA fees come on top, and we break those down in our piece on Amazon FBA fees. In your own store, the same order loses the payment provider's fee, usually a fraction of the marketplace rate, plus the shipping you were paying anyway.

The same arithmetic undoes the popular justification “we will build a store and stop paying commission”. You do not stop paying it, you swap it for the cost of acquiring traffic. That trade pays off if, and only if, the customer comes back.

Customer data: what is allowed and what is not

The most common misunderstanding in this conversation sounds like this: “we have thousands of customers on Amazon, we will email them about the new store”. You will not. The agreement with Amazon (Amazon Services Europe Business Solutions Agreement, clause 11.2, version of 18 September 2025) states that customer personal data, contact information included, is processed to fulfil orders, provide customer service and comply with applicable law, and only for those purposes.

The practical conclusion: your Amazon sales history is not a marketing list. It is proof that the product sells, and that is all. A list you may write to only starts to exist in your own channel, with its own marketing consent, on your terms. That is usually the real reason brands build a store, far more often than fear of a suspension.

Signals that this is not the moment

• One product with no repeat purchase. The customer buys once and has no reason to return. The break-even above is then out of reach.
• A margin that cannot carry advertising. If a few per cent survive commission and logistics, there is nothing to fund traffic with. How to calculate it is in our piece on Amazon profit margin.
• No customer service on your side. Under FBA, returns and questions go to Amazon. In your own store that is your phone and your inbox, seven days a week.
• Product data that lives only in Seller Central. A store built on manually retyped descriptions drifts away from Amazon within a quarter. The starting point is covered in our piece on product data for Amazon.
• “Because everyone has one” as the reason. That is not a business case, it is a fixed cost for years ahead.

One catalogue, two channels: what breaks first

Once the decision is made, the store stops being the problem and running two channels from one catalogue becomes one. In our practice, three things break fastest.

• Stock levels. Two channels selling from one warehouse without integration end in overselling, and on Amazon overselling hits account health. Wiring this into Seller Central is covered in our piece on connecting a store to Amazon.
• Prices. A lower price in your own store looks obvious, since you pay no commission. Before you set it, check the current Amazon pricing policy in your account, because prices in other channels can be taken into account when an offer is assessed.
• Product content. Descriptions written twice always drift apart. One source of truth for the catalogue, feeding both Amazon listings and store pages, is the difference between running two channels and running two companies.

Also assume the customer will bring Amazon expectations into your store: fast delivery, simple returns, no questions asked. That is a real operating cost, not an interface detail.

How this works on our side

We start with the arithmetic, not with a store design. We look at repeat purchase, margin after fees and whether the brand already has demand of its own. If the numbers do not work, we say so and stay with growing the Amazon channel, because the same money works faster there. If they do, we run both sides: marketplace sales under account management and the build under eCommerce platform development.

The technology side, the platform itself, is handled in our group by our sister brand Seedlight, for example in its D2C eCommerce service, or B2B pricing and quotes for wholesale. We mention it not to close the article with an advert, but because this is exactly the decision where brands ask us who should build it.

Let us run the numbers on your case →

FAQ: an own store alongside selling on Amazon

Does an own store protect me from an Amazon account suspension?

Not in the way people expect. Amazon customers will not move to your store automatically, because most do not know it exists. A store gives you a second leg to stand on only if you built it earlier and it already has traffic of its own.

Can I email customers who bought from me on Amazon?

No. The agreement with Amazon limits processing of buyer data, contact information included, to fulfilling orders, customer service and legal obligations. A marketing list is something you build in your own channel, with separate consent.

How many orders does a customer have to place for the store to pay off?

As many as your customer acquisition cost divided by the difference between what is left from an order in your store and what is left after the Amazon fee. For a typical basket that lands at several orders, so repeat purchase decides, not sales volume.

Can I price lower in my own store than on Amazon?

Technically yes, but check the current Amazon pricing policy in your account first. Prices in other channels can be taken into account when an offer is assessed, and losing visibility costs more than the commission you save.

Where do I start if the numbers work?

With the catalogue, not the look of the store. One source of product data feeding both Amazon listings and store pages, plus stock integration. The rest is simpler than it looks.

Sources

• Amazon, fees on amazon.pl (referral fee by category, charged on the full amount including shipping), accessed 27 October 2026
• Amazon, Amazon Services Europe Business Solutions Agreement, clause 11.2, version of 18 September 2025