Amazon or Shopify: Which Channel Actually Leaves More Profit?

Look at an Amazon settlement and it is easy to wonder how so many businesses have managed to help themselves to your sale. Referral fees, fulfilment, storage, advertising. You supplied the product, but there appears to be quite a queue at the till.

Then someone suggests Shopify. Your own website, more control over the brand and a much smaller payment fee. Surely that leaves more profit?

It can. But opening a Shopify store does not bring the customers with it. If you save £8 in marketplace charges and spend another £12 persuading someone to visit your website, the saving has become rather expensive.

My view is that both channels can earn their place. The decision should come from what each order contributes, how reliably you can generate those orders and what running the channel costs the business. A website is not automatically the more profitable option, and Amazon's fee statement is not enough reason to abandon it.

We will compare the same £60 product on both channels, then look at the decisions that change the result. The useful outcome is knowing where to put the next pound of investment.

You are buying two different routes to the customer

Amazon brings a marketplace where people already search for products, compare options and place orders. You still need a competitive offer, convincing listings and often advertising, but you are entering an existing shopping environment.

Shopify provides the infrastructure for your own store. You control much more of the presentation, merchandising and customer journey. Getting suitable people to the shop is your responsibility, whether through advertising, search, social media, referrals or an audience you have already built.

That difference matters. A brand with an engaged following starts its website from a different position from one whose customers have only ever found it through Amazon search. Copying the product listing onto a website does not copy the demand.

I would also separate the sales channel from fulfilment. Amazon orders can be fulfilled by Amazon through FBA or dispatched by the seller. Shopify orders might be packed by your team or an external fulfilment provider. Compare the arrangements you would actually use, rather than assuming every Amazon order and every website order has the same cost structure.

Put the same product through both calculations

Take a UK VAT-registered brand selling a standard-rated product for £60, including delivery and 20% VAT. The revenue excluding VAT is £50. Its landed product cost is £15, including the relevant inbound freight and duty, excluding recoverable VAT.

The following figures illustrate a retained, single-item order. They are assumptions for a comparison, not a quote for your product. We use Amazon FBA on one side and third-party fulfilment on the Shopify side. Costs exclude recoverable VAT, and any non-recoverable tax belongs in the cost.

Shopify saves £7.55 on the referral or payment fee, but delivery costs £1 more and advertising costs £8 more. Under these assumptions, Amazon leaves £1.45 more per order.

The fee comparison favours Shopify. The complete order comparison favours Amazon. That is why I want the whole calculation before deciding which channel is expensive.

The Amazon example assumes a category with a 15% referral fee. Importantly, we calculate that on the £60 customer payment: £9, rather than 15% of the £50 net revenue. Your category, product dimensions, weight and fulfilment arrangements determine the actual charges.

The Shopify payment assumption is 2% of £60 plus 25p, giving £1.45. This matches the standard online card rate displayed for Basic on Shopify’s UK pricing page. Plans and payment methods can change the rate; additional third-party transaction fees may apply with another payment provider. Check your own settings before using the figure.

The fulfilment and other allowances are illustrative, not published platform tariffs. This table also excludes monthly subscriptions, agency retainers, website development and other overheads. Those come next. What remains here is contribution, not the owner's take-home profit.

Work out how much acquisition cost the fee saving can support

Our Shopify order leaves £26.55 before advertising. To match Amazon's £14 contribution after advertising, Shopify can spend £12.55 per retained order on advertising: £26.55 less £14.

That gives the founder a useful decision point. At £10 advertising per order, Shopify contributes £16.55 and wins this comparison. At £18, it contributes £8.55 and falls well behind. If Shopify requires more channel overheads, its advertising allowance needs to be lower again to deliver the same operating result.

Here we are allocating advertising across retained orders, including orders that arrive without a paid click. That is a blended order measure. Customer acquisition cost answers a different question: how much you spend to acquire each new customer. Keep both visible when repeat orders are significant.

My target for a scalable own-brand model remains paid advertising around 10–20% of net revenue. On our £50 net sale, that is £5–£10. In the example, Amazon is at 12% and Shopify at 28%, so I would challenge what needs to change on the website side.

An established customer base may support a more expensive first acquisition, provided repeat purchasing and cash payback justify it. But a high acquisition cost needs a commercial explanation. Saving marketplace fees is not one on its own.

Give repeat customers a value you can actually evidence

A website can become much more attractive when customers return without needing the original acquisition spend again. Relevant email campaigns, replenishment reminders, subscriptions or useful new products can help build that relationship.

The opportunity is real. The forecast still needs evidence.

Suppose 100 new Shopify customers each place one order under our example assumptions, with £14 advertising cost per customer. Their first orders generate £1,255 contribution after advertising.

If 30 of them place one further identical order within 90 days, and retention activity costs £2 per repeat order, each repeat contributes £24.55: £26.55 before marketing less £2. The repeat orders add £736.50, taking the group's contribution to £1,991.50 before fixed overheads.

That is a more promising picture than judging the first orders alone. It also depends on those 30 customers returning, the price holding and the costs staying comparable. It is an illustrative scenario, not a repeat-purchase benchmark.

Track customers acquired in the same month and measure their actual contribution over 30, 60 and 90 days, or a period suited to the product. Compare marketing spend with the cash contribution earned back. Lifetime revenue does not pay an invoice due next Friday.

Amazon customers can also buy again. Include the repeat orders and the associated fees in that channel's results. Do not assume every repeat order belongs to Shopify's side of the argument, or assume marketplace customer information is available for unrestricted direct marketing.

A better basket can matter more than a cheaper transaction

Shopify gives you room to design bundles, complementary offers and a buying experience around the range. The financial benefit comes when customers buy more useful product without acquisition and fulfilment costs increasing at the same rate.

For example, suppose a Shopify customer adds an accessory priced at £18 including VAT. That adds £15 net revenue. If its landed cost is £4, additional payment processing is 36p and additional packing and delivery cost is £1, it adds £9.64 contribution before any extra returns or marketing costs. The payment's fixed 25p charge has already been incurred on the original basket.

That is meaningful. But if a second parcel is needed, or a discount across the whole basket gives away most of the gain, the result changes. Measure contribution per order alongside average order value. A higher basket total is only useful if the extra products earn their space in it.

Amazon can also support multipacks and bundles, subject to its requirements. Compare the resulting product fees and fulfilment costs. The opportunity belongs to the offer and its economics as much as to the platform.

Monthly costs can overturn the result

A channel that contributes more per order can still contribute less to the business once its specific overheads are included.

Suppose each channel generates 500 retained orders using the first table. Amazon produces £7,000 contribution after advertising; Shopify produces £6,275. If channel-specific monthly overheads are £300 for Amazon and £900 for Shopify, their contributions towards shared business overheads become £6,700 and £5,375 respectively.

Those are illustrative budgets, not standard running costs. The extra website spending might cover apps, maintenance, content or customer support. Agency and creative costs need including somewhere too. Add them once, whether allocated to orders or shown as channel overheads.

At this volume, the £600 overhead difference is £1.20 per order. Shopify would therefore need advertising at £11.35 per order to match Amazon's result after these channel costs, assuming the other figures stay unchanged.

There is a practical lesson here for a low-volume store. An extra £200 a month of apps costs £2 per order at 100 orders, but 20p at 1,000. A small subscription can be a surprisingly enthusiastic participant in a small business's margin.

Separate costs that disappear if a channel closes from shared costs that remain. Allocating half the office rent to Shopify might help reporting, but closing Shopify will not automatically halve the rent. Equally, staff time is not free because the same person already works for you. Record the workload and the point at which it requires another hire.

More margin per order does not always mean more profit

Imagine Shopify earns £18 per retained order and Amazon earns £12. Looking only at those figures, Shopify appears the obvious place to focus.

But 300 Shopify orders produce £5,400 contribution, while 1,000 Amazon orders produce £12,000. Both can be valuable. What matters next is how many additional orders each can generate and what those additional orders will cost.

Do not multiply today's average result by tomorrow's hoped-for volume. Expanding advertising into colder audiences may make the next 300 customers more expensive. A larger operation may also require extra support, storage or a different fulfilment arrangement.

Before increasing a budget, agree the additional contribution you expect and the spending limit. Review the result after refunds and the relevant costs have had time to emerge. If the next £2,000 of advertising produces only £1,600 contribution before that advertising, the increase has lost £400 before any extra overheads. A respectable historical average does not rescue it.

Include the stock and cash needed to run both channels

Two profitable channels can still put pressure on the same bank account.

Stock may sit in an Amazon fulfilment centre while the website warehouse runs short. Replenishing both locations can require a larger order, more safety stock or extra transfers. If adding the channel ties up another £10,000 in inventory, include that commitment when assessing its return and funding needs.

Forecast when receipts reach the bank, rather than treating an order notification as available cash. Payment processing, settlement arrangements, account reserves and refunds can affect that timing. Use your own provider and marketplace account information in the forecast.

Meanwhile, advertising, VAT and supplier payments keep their dates. A channel making a respectable profit but needing months of funding may be harder to scale than one delivering a smaller, faster return. That is a purchasing and cash decision as well as a margin decision.

Make the accounts answer the comparison properly

A platform payout is not the sales figure. It may already have fees, refunds and other deductions removed, and its timing may differ from the underlying orders. Comparing Amazon's net bank receipt with Shopify's gross order total will produce a very confident answer to the wrong question.

At COPA, I want the monthly information to show net revenue, landed product costs, selling fees, fulfilment, advertising, returns and channel overheads on a consistent basis. Tools such as Link My Books can help bring settlement information into Xero, but stock costs and the wider marketing bill still need proper attention.

Start with the same products where possible, then assess each channel's actual product mix. A website selling mostly high-margin accessories and an Amazon account selling the main product are not directly comparable until you understand that mix.

Record sales after discounts and refunds, excluding VAT. Our first table illustrates a retained order with an allowance for additional unrecovered returns costs. Across the full business, reduce revenue for refunded orders and include their actual unrecovered costs; do not deduct the same refund again through a second blanket allowance.

Useful monthly comparisons include contribution per retained order, total channel contribution, advertising as a percentage of net revenue, new-customer acquisition cost, repeat purchasing and the cash tied up in stock. Tie the totals back to the accounts. Five reconciled measures will tell you more than twenty impressive charts built on different definitions.

Choose a role for each channel

A strong Amazon operation does not need replacing just because owning a website sounds more sophisticated. A profitable Shopify store does not need pushing onto Amazon without checking the category fees, pricing and operational demands first.

Using both can broaden reach and reduce dependence on one source of orders. It can also create duplicate work without much additional business if customers simply move between the two. Test whether the new channel is adding customers and contribution across the whole brand.

I would start with a manageable range, a defined budget and a clear reason for the test. Price the products properly, reconcile the transactions and review actual contribution and cash use before extending the commitment. Keep launch costs separate so you can see both the ongoing economics and how long the original investment will take to recover.

My answer to Amazon or Shopify is to give each channel a job and make it earn the resources it uses. Amazon may bring profitable demand at scale. Shopify may support stronger baskets and repeat relationships. Your figures should establish which opportunity is working for your business.

At COPA Accounting, we help ecommerce founders connect marketplace activity, stock, marketing and cash to the accounts. If both platforms say sales are growing but you cannot tell which one is paying for the growth, that is the first calculation I would want us to fix.