A shopper types your brand name into Amazon, clicks your ad and buys. Amazon counts a win. That shopper was already coming to you, and on your own site you would have kept more of the sale.

Your Amazon ads report a ROAS of 5: £4,000 spent, £20,000 of sales credited. How many of those sales needed the ad?

Where the 5 comes from

Say your Amazon margin, after referral and fulfilment fees, is 28%. Break-even ROAS is 1 ÷ 0.28, about 3.6 (the return an ad needs to cover the cost of each sale; work out yours in Easier Ads Budgeting). The 5 looks safe. Now remove the credit for brand searches:

Sales credited£20,000
−
From people searching your brand£8,000
=
Sales from shoppers who did not know you£12,000

Those searchers were heading to you anyway. £12,000 on £4,000 is a ROAS of 3, under the 3.6 break-even. The campaign that looked best loses money on its real work.

The rule

Spend Amazon ad money on shoppers who do not know your brand. Leave brand searches to your own site.

Step 1: know who is searching

Visitors to your store know you. They buy at full margin and join your email list.

Someone typing “collagen supplement chocolate flavour” on Amazon has not heard of you. They compare 7 brands on reviews, price and photos.

Step 2: fence off your brand

Run a small brand defence campaign with modest bids, so competitors cannot take your name. You win most of those searches cheaply. Do not push more in.

Step 3: put the rest on category demand

A starting split of your Amazon ad budget, adjusted as data comes in:

  • Auto campaign, 20%. Amazon finds terms you had not thought of. Read the search term report weekly.
  • Category exact match, 50%. Your proven category terms, bids set from your Amazon break-even ROAS.
  • Competitor product pages, 20%. Appear where shoppers compare.
  • Brand defence, 10%.

Step 4: give people a reason to buy direct

Matching Amazon’s price on your site wins nothing. A small premium with something added does: an exclusive variant, a free gift, early access to launches. Keep Amazon pricing competitive with other sellers there.

Step 5: check Amazon adds sales

Each month, track:

  • New-to-brand share of Amazon orders, in your Amazon Ads reports. If it keeps falling, more spend reaches people who know you.
  • How many Amazon buyers later buy on your site. Add “how did you first hear about us?” to your post-purchase survey and follow the Amazon answers for 12 months.
  • Blended cost of a new customer, Amazon spend included. If it gets worse as Amazon spend grows, check new-to-brand share before adding more.

If Amazon and your store are pulling against each other, book a free call and we’ll read it with you. Or see how our Amazon ads team works.

Quick answers

Do Amazon ads cannibalise Shopify sales?

Sometimes, and it is mostly avoidable. It happens when a shopper who would have bought on your own site buys on Amazon instead, where fees take a bigger share of the sale. The risk is highest on your own brand name. Spend Amazon ads on category searches, where the shopper does not know you yet.

Should Shopify brands use Amazon at all?

For most physical product brands, yes. Amazon works best as a way to be found by people who have not heard of you. Some of them search your name later and buy direct, at a higher margin.

Should I bid on my own brand name on Amazon?

Only a small defence campaign, so a competitor cannot sit on your name. Heavy brand bidding pays Amazon to reach shoppers who were already heading to you.

What is a good target ROAS for Amazon ads?

Work it out from your Amazon margin, not your website margin. Amazon's referral and fulfilment fees take a bigger share. Break-even ROAS is 1 divided by your margin after those fees, so a 28% margin gives about 3.6.

Found this useful? Book a free strategy call