Should I sell on Amazon and Shopify?

The decision is arithmetic before it is strategy. Take one representative product and build the contribution for each channel, then look at what is left.

For the marketplace: price, minus cost of goods, minus the referral fee, minus fulfilment and storage if you use the platform's warehouse, minus an allowance for returns and for the units that come back unsellable. What remains is what you have to work with, before any advertising at all.

For your own store: price, minus cost of goods, minus payment processing, minus shipping and packing, minus your returns allowance, minus the platform and app subscriptions divided across your orders. Then subtract what it costs to bring the visitor, because nobody walks past your store by accident.

Now the useful comparison. The marketplace has a high fixed cost and free footfall. Your own store has a lower fixed cost and paid footfall. Whichever ends up with more contribution per order at your real volumes is the channel with the better economics today, and the answer is often that they are close enough that both are worth running.

Where the two diverge is what happens next. On your own store the second order costs almost nothing, because you have the email address. On the marketplace you usually pay the same acquisition price again. Run the arithmetic across two or three orders per customer and the picture changes, which is why the answer for a consumable is different from the answer for something bought once.

It gives you demand that already exists. People arrive with a purchase in mind, payment details stored, delivery expectations already set by the platform, and a willingness to buy from a brand they have never heard of because the platform stands behind the transaction. For a new brand that is worth a great deal, and it is very hard to reproduce.

It also gives you a fast read on demand. Listings tell you which variants sell, what price the category clears at, and which product claims matter, all without building a marketing programme first.

What it takes is a set of things you may not miss until later. You do not get the customer: no email address you can use freely, limited ability to market to buyers, and no way to build an audience from their behaviour. You compete on a page that also shows alternatives, so price and reviews carry more weight than brand. You are exposed to listing problems, other sellers on your own listing, suppressed offers and category rule changes, any of which can stop revenue overnight without warning.

And you inherit a review dependency. The listing lives or dies on its rating, so one bad batch or one packaging failure costs more than the units involved.

None of that argues against selling there. It argues for treating it as a channel with a landlord rather than as your business. Keep a route to your own customers, and keep the brand assets, the content and the customer list somewhere you control.

Your own store gives you margin, data and control. You set the price, the bundle, the delivery promise, the upsell and the follow up. You can offer a subscription, a size that would confuse a marketplace listing, or a kit that only makes sense as a set. You keep the email address and the order history, which is what makes the second and third order cheap.

It also gives you a brand people can actually visit. Search results, content, comparisons and social all point somewhere that is yours, and none of that work is at risk from a policy change on a platform you do not own.

The bill is traffic. Every visitor either costs money or costs the time it takes to earn organic visibility. That is the single biggest difference between the channels, and it is why a store with a good product and no traffic plan reports that its own site is dead while its marketplace listing sells.

The second part of the bill is conversion. On your own site you have to do the reassurance work the marketplace does for free: delivery dates, returns policy, reviews, payment options, and enough detail that nobody leaves to check the price elsewhere.

The third part is running it. Photography, copy, page speed, checkout, apps, tax settings and support all sit with you. That is a real workload, and it is the reason to add the second channel deliberately rather than in a week when things are already busy.

If you already sell on a marketplace and are considering your own store, start narrow. Take your best sellers, build proper pages, and use email capture, subscriptions and bundles to earn the margin the marketplace cannot. Do not try to move the whole catalogue at once. The aim in year one is a customer list and a repeat rate, not matching marketplace volume.

If you already have your own store and are considering a marketplace, also start narrow. List a small set of products that survive a price comparison, with correct identifiers, real photography and stock you can hold. A thin catalogue with strong listings does better than a full catalogue that runs out.

Then budget for the operational overhead honestly. Inventory has to be allocated between channels, and an oversell is a suspended listing on one side and a cancelled order on the other. Product data has to stay in step: titles, images, identifiers and prices. Returns arrive through two different processes with two different windows. Support answers questions from two places with different tone rules. Reporting has to be stitched together, because neither system sees the other.

The honest test of readiness is whether one person can own that overhead without dropping the channel that already pays the bills. If the answer is no, the second channel is a next quarter project with a hire or a system behind it, not a weekend launch.

One rule to hold across both: whatever you build, keep the accounts, the domains, the ad accounts and the customer data in your own name so the channel mix stays a choice you can change.