Amazon Ads ROAS explained: attributed sales for each advertising dollar
Return on ad spend divides attributed sales by advertising spend. A ROAS of 4 means the report contains four dollars of attributed sales for each dollar of ads. It does not mean the author earned four dollars or made a three-dollar profit.
This page translates the multiplier into a funding decision. Use the ACOS guide for spending as a sales percentage and the CTR guide for the click stage. The three metrics describe different questions, not three independent proofs of a successful campaign.
Use the report's sales value
ROAS = attributed sales / advertising spend. A fictional $120 sales on $30 spend gives ROAS 4, also written 4:1. Amazon's ROAS guide defines the relationship. For books, keep the distinction between retailer sales and author earnings explicit.
Handle ratios and missing values correctly
With spend in column B and sales in column C, =IF(B2>0,C2/B2,NA()) returns a multiplier. Do not format it as a percentage unless you deliberately want a different presentation. If spend is positive and attributed sales are zero, ROAS is 0. If spend is zero, the ratio is undefined, not a meaningful infinite success score.
For positive spend and positive sales, ROAS = 100 / ACOS expressed as a percentage. Thus 25% ACOS corresponds to ROAS 4. The identity requires the same dates, currency and sales scope; mismatched exports need reconciliation, not a mathematical workaround.
1. Translate retail sales into the author's contribution
Estimate the contribution associated with those sales using the applicable editions and costs. KDP's eBook royalty guidance explains why the payout basis depends on the option, territory and delivery costs. Print and other editions require their own calculation.
Higher ROAS can still mean a worse result
| Campaign | Attributed sales / ROAS | Assumed contribution before ads | Contribution after ads |
|---|---|---|---|
| Edition A | $120 / 4 | 20% x $120 = $24 | $24 - $30 = -$6 |
| Edition B | $60 / 2 | 60% x $60 = $36 | $36 - $30 = $6 |
The 20% and 60% ratios are teaching assumptions, not standard royalties for named formats. Campaign A wins a ROAS leaderboard and loses money under its stated contribution model. Campaign B has a lower multiplier but leaves a positive contribution before other overhead.
Amazon's advertising-math guide likewise ties break-even return to margin. For an author, substitute a transparent contribution model. Do not turn a generic retail benchmark into an earnings promise for a low-priced book.
2. Calculate break-even and a more useful funding target
Break-even requires a consistent sales basis
If contribution is a stable fraction m of the same sales value used in the ad report, break-even ROAS is 1 / m, before costs outside that contribution. A fictional 40% ratio gives 1 / 0.40 = 2.5. At that return, contribution just covers advertising.
Use 0.40, not 40, in the reciprocal calculation. If contribution varies materially across editions, calculate the mix or inspect them separately. The result is only as useful as the underlying royalty, cost and sales assumptions.
Retained contribution changes the target
Suppose an edition has a matching $5 sales value and $3 contribution per purchase, but you want to retain $1 after ads. Only $2 is available for acquisition, so target ROAS is $5 / $2 = 2.5. Its simplified ad-only break-even is $5 / $3, approximately 1.67. These targets serve different objectives.
Other expenses can raise the required target further. Record whether management, creative production or overhead is included. Do not call contribution after advertising complete net profit if those costs remain outside the model.
3. Combine sales and spend, not the displayed multipliers
Imagine one report row with $60 sales on $10 spend, ROAS 6, and another with $180 sales on $90 spend, ROAS 2. Combined ROAS is ($60 + $180) / ($10 + $90) = 2.4. The simple average of 6 and 2 is 4, which overweights the smaller row.
For compatible non-overlapping spreadsheet rows, use =IF(SUM(B2:B3)>0,SUM(C2:C3)/SUM(B2:B3),NA()). Do not sum a campaign total with its own ad-group or target breakdown. Retain the detail: a valid aggregate can still conceal an unaffordable edition or traffic source.
| Comparison field | Record | Why it matters |
|---|---|---|
| Edition and currency | Format, territory, matching money basis | Sales and contribution can differ materially |
| Observation period | Dates and report extraction time | Fresh and matured results are not equivalent |
| Attribution | Supported interactions and lookback | Click-only and broader conversions are different scopes |
| Audience job | Existing interest, discovery or another stated goal | Different contexts can produce different apparent efficiency |
| Costs | Included and excluded expenses | Retail ROAS is not full-business return |
4. Align attribution before claiming a change in return
Amazon's attribution help distinguishes campaign and account rules, including a 14-day click window for Sponsored Products authors and no book brand halo. Check the actual report definition; do not apply every seller or display-campaign rule to an author campaign.
Attributed return is not proven incremental return. A reader may already know the author or encounter an email, discount or other promotion during the same period. Keep those events in the log instead of claiming that the ad caused every attributed purchase.
Allow results to mature without authorizing unlimited spend
Export a later version of recent periods after the applicable window and processing time. Preserve earlier downloads so changes can be understood. A campaign can be paused while older interactions gain qualifying conversions. Respect the cash allowance while waiting.
5. Keep Kindle Unlimited and read-through estimates in a separate model
Amazon's metrics help describes estimated attributed KENP reading and royalties where available. Identify reading as reading, not paid retail sales. Do not add all account-wide KU income to one campaign or silently alter a standard report's numerator.
For a separate fictional worksheet, suppose $30 spend produces $40 attributed paid sales from eight $5 purchases. Paid-sales ROAS is 1.33, rounded. At $3 assumed contribution per purchase, paid contribution is $24. Add $10 of eligible, non-duplicated estimated attributed KU royalties and the estimated combined contribution becomes $34, leaving $4 after ads before other expenses.
That last calculation is an estimated contribution model, not a relabeled retail-sales ROAS. Reconcile reading estimates with publishing reports and preserve attribution limitations. Do not also count the same $10 elsewhere in the return.
Series read-through can be investigated separately with a defined cohort, period and cost basis. Do not assign every later purchase of another title to the first ad or use hoped-for lifetime value to excuse an unlimited current loss.
6. Evaluate additional spending, not just the historical average
| Scenario | Spend / sales / ROAS | Contribution after ads |
|---|---|---|
| Initial observation | $30 / $60 / 2 | $36 - $30 = $6 |
| Expanded observation | $45 / $75 / approximately 1.67 | $45 - $45 = $0 |
The second scenario adds $15 spending and $15 sales value, only $9 additional contribution under the stated model. The extra activity does not pay for itself. This is arithmetic for a hypothetical comparison, not a causal claim from a real before/after experiment.
Higher authorization can reach a different traffic mix and change the average result. Use a bounded change and record concurrent factors. A historically acceptable ratio is not proof that doubling a budget will double profitable orders.
7. End the review with a funding decision
- Verify positive spend and the correct sales definition.
- Align currency, period, edition and report grain.
- Translate sales into the stated contribution model.
- Compare against break-even and the desired retained amount.
- Separate paid sales, estimated reading and speculative read-through.
- Choose renew, bounded follow-up, redesign or pause with a cash limit.
Use ACOS when a percentage is easier to interpret, and CTR when the problem appears before the purchase stage. Return to the Ads hub for specialist targeting and optimization guidance.
Sources and scope
Official Amazon Ads and KDP documentation checked: 14 October 2026. Dollar amounts, ratios and scenarios are fictional teaching examples. They describe attributed results and explicitly stated contribution assumptions, not guaranteed incremental income, universal benchmarks or complete business ROI.