Ready to take your Amazon Vendor data ops to the next level?
Request access today and let InsightLeap help grow your business!
June 11, 2023 by InsightLeap
Amazon ROI = (Profit - Investment) / Investment × 100. Profit is what your Amazon sales return after selling costs like referral and fulfillment fees, advertising, and any shipping you cover, and investment is what you spent getting the goods into Amazon's network. The arithmetic is easy. The number goes wrong in the ledger behind it, because every cost line you leave out inflates the result. What follows is the calculation with a worked example, then the cost ledger it depends on, then five levers that move the ratio itself.
Two numbers drive it:
Amazon ROI = (Profit - Investment) / Investment × 100
Profit is total revenue from your Amazon sales minus the costs of selling: Amazon fees (referral fees, plus fulfillment fees if Amazon ships your orders), advertising, and any shipping you cover. Investment is the total you spent to make the goods sellable in the first place: sourcing or manufacturing, packaging, and inbound freight to Amazon's fulfillment centers. Count every dollar exactly once, though, because a cost sitting on the investment line should not come off the profit line as well.
Take a hypothetical with round numbers. Say you buy 200 units at $9 each and pay $200 in inbound freight, so $2,000 is invested. All 200 sell at $25, bringing in $5,000 of revenue. Referral and fulfillment fees take $1,250, advertising takes $500, and the shipping you cover takes $250, which leaves $3,000 of profit. ROI = ($3,000 - $2,000) / $2,000 × 100 = 50%. Your own fee lines will differ by category and program, but the shape of the calculation doesn't change.
ROI and profit margin are easy to conflate, and they answer different questions. The example above earns 50% on the $2,000 put in, while the same $1,000 of net gain is a 20% margin on $5,000 of revenue. Margin tells you what each dollar of sales keeps, and ROI tells you how hard the cash you tied up worked, which is the more useful question when you're deciding where the next purchase order goes.
When an ROI comes out suspiciously good, the usual culprit is a ledger with lines missing. All of these belong in it:
Selling 1P adds lines that most ROI guides skip. On Vendor Central, Amazon buys from you wholesale, and your revenue line is the cost price you agreed to, so the margin math starts from the PO. Then the deductions come off: chargebacks and shortage claims are cost lines like any other, and an ROI computed before them is higher than the cash that actually lands. Our overview of Amazon Vendor Services covers how the PO relationship works.
There is no universal benchmark, and Amazon's guidance on marketing ROI says as much: establish your own baseline and track progress against it. What counts as strong depends on your category economics, your program mix, and what the same cash could earn somewhere else. Compute the honest version from the ledger above and treat it as the baseline you judge next quarter against.
Every lever below works on the ratio: more profit out of the same investment, or the same profit on less cash tied up. Growing revenue on its own can leave the ratio exactly where it started.
If your current ROI came off incomplete costs, you can't tell which levers are working, because every fee line you left out has been inflating the number all along. Re-run last quarter through the full ledger before you touch strategy. It's an afternoon of work, and it gives the four levers below a number worth measuring against.
Chargebacks and shortage claims sit in the ledger as cost lines, and some of them should never have been taken. Reconcile them against your own records and dispute the invalid ones. For 1P vendors this is usually the cheapest ROI improvement on the board: it needs no ad budget and no price change, only the hours to work through the remittances, and every recovered dollar goes straight onto the profit line.
Two formulas govern ad efficiency, and both come from Amazon's guide: ROAS = ad revenue / ad spend, and ACoS = (ad spend / ad revenue) × 100, where a lower ACoS means less spend for the sales it generated. Neither means much until you set it against your margin.
Gross profit is revenue minus the cost of goods sold, and the gross margin percentage that falls out of that is your break-even ACoS. Amazon's ACoS guide makes the link explicit: to stay profitable, ACoS has to stay below your profit margin, so a product with a 30% gross margin breaks even at an ACoS of 30. A campaign running above its break-even is buying revenue at a loss even when the dashboard looks healthy. Work out the break-even for each product, judge campaigns against that number instead of an account-wide target, and move spend toward the ones that clear it. For a longer conversation about maximizing returns on Amazon advertising, see our interview with Kenshoo's Nich Weinheimer.
A sale that arrives without a paid click carries no ad cost, so every point of organic share lifts ROI directly. The work here is listing quality and reviews: complete product content, images that answer the questions shoppers ask before they buy, and a review base someone on the team is responsible for. Track what share of your orders arrives organically and treat moving that number as a project in its own right, separate from growing total sales.
An account-level ROI is an average, and averages hide the products dragging them down. Run the ratio per ASIN with the same ledger and the spread shows up: usually a handful of products carrying the account, and a tail where fees, freight, and ad spend eat more than the product returns. Some of that tail has an obvious fix. The rest is structural, and cutting those ASINs frees the budget to go behind products that already clear your baseline.
An ROI number is a snapshot, and it goes stale: fees change, ad auctions move, and a quarter's worth of deductions can land in a single week, so the figure you computed last quarter stops describing the business you have now. Recompute on a schedule instead, weekly if your catalog is ad-heavy and monthly at the very least, and pull from the same feeds every time so that a change in the number means a change in the business rather than a change in method. For 1P vendors the raw material sits in Vendor Central, where Amazon Retail Analytics covers sales performance, inventory, and customer behavior.
The failure mode is almost always manual reporting: someone rebuilds the spreadsheet by hand for a month or two, then a busy week arrives and the cadence dies. Skipping that regular read has a real cost.
Every number on this page can be produced by hand from downloaded reports. What kills the cadence is the hours that takes every week. InsightLeap automates that reporting layer for brands and agencies, pulling and assembling Vendor Central reports on a schedule so the ledger and the trend stay current without the spreadsheet work.
Customers mostly describe the difference in time. Navitas Organics says automating product content audits alone saved more than 50% of their time a month (their story). General Tools reports Amazon sales up 31%, with the caveat that InsightLeap is not solely responsible for that growth, and a couple of team hours saved every day (the case study).
Request access today and let InsightLeap help grow your business!