Is revenue growth leaving any profit?
A reported return on ad spend is a ratio, not a profit statement. A campaign generating $3 of attributed revenue for every $1 of advertising can be profitable, barely profitable or loss-making depending on the rest of its costs. Begin with what is left before advertising rather than choosing a ROAS target in isolation.
This single-order example applies an arithmetic model to ecommerce advertising. It does not predict a Google auction, prescribe an automated-bidding setting or assume all attributed sales were incremental. Reported conversion value must use the same revenue definition as the model for a direct comparison to make sense.
Define revenue and costs consistently
Let P be order revenue, C non-ad costs per order, m target margin as a decimal and v conversion rate as a decimal. Here, P is $80 and C is $50. Non-ad costs include goods, delivery, packaging, processing, expected return losses and any chosen overhead allocation.
A revenue figure containing collected tax should not be compared with a price excluding it. Similarly, do not deduct a refund allowance if those same refunds have already reduced revenue. Multi-currency results need a common reporting currency.
Contribution before ads = P − C = $30.
Break-even CPA = $30.
Break-even ROAS = $80 ÷ $30 = 2.67×.
This threshold leaves zero profit after the entered costs. It is not the return required to meet a positive business target.
Build the target-profit threshold
A 15% margin means keeping $12 on an $80 order. That leaves $18 for advertising, so the target ROAS is $80 ÷ $18 = 4.44×.
Target CPA = P − C − (P × m).
Target ROAS = P ÷ target CPA, provided target CPA is positive.
If the allowance is zero, only zero acquisition spend can meet the target in this simplified model. There is no finite positive-spend ROAS target. If the allowance is negative, even a freely acquired order misses the target. Display a warning rather than presenting a negative or infinite number as a recommended return.
Compare actual returns with the target
Keep price at $80 and non-ad costs at $50. Advertising cost per order is price divided by ROAS.
| ROAS | Implied advertising cost | Modeled profit per order | Margin |
|---|---|---|---|
| 2.00× | $40.00 | −$10.00 | −12.50% |
| 3.00× | $26.67 | $3.33 | 4.17% |
| 4.00× | $20.00 | $10.00 | 12.50% |
| 5.00× | $16.00 | $14.00 | 17.50% |
At 3× the campaign covers the stated costs but falls far short of a 15% profit margin. Improving the displayed ratio only helps interpretation when the revenue, attribution and cost assumptions remain comparable.
Translate CPA into a click-cost allowance
CPC = CPA × conversion rate. With an $18 target CPA:
| Assumed conversion rate | Corresponding target CPC |
|---|---|
| 1% | $0.18 |
| 2% | $0.36 |
| 3% | $0.54 |
These are planning outputs, not bids guaranteed to acquire traffic. A small conversion sample can make the assumed rate unreliable. Separate device, product and campaign differences before treating one average as representative of everything.
Stress-test costs rather than trusting one average
At the same $80 price and $12 target profit, $45 non-ad costs permit a $23 CPA and require about 3.48× ROAS. At $58 non-ad costs, the allowance shrinks to $10 and the target rises to 8×. A $13 cost difference has changed the advertising requirement dramatically.
For a varied catalog, evaluate the products generating the attributed sales. Revenue-weighted contribution is more useful than an unweighted average of SKU margins. Check for duplicated conversions and delayed refunds before calling the campaign profitable.
Apply the result to a real decision
Use the Google Shopping calculator for the full scenario and Break-Even ROAS for a simpler threshold. The returns guide helps when refunded orders materially change contribution.
Before increasing spend, compare the implied CPA with observed acquisition costs. Check that conversion value is sales revenue rather than an arbitrary assigned number: a page-view conversion assigned $80 is not an $80 sale. These worked examples are original arithmetic scenarios, not evidence of future campaign performance.
Try different assumptions
Defaults reproduce the hypothetical example. Edit the inputs and recalculate. This is a separate teaching scenario, not a live platform quote.
Another example: why 3× ROAS can still lose money
This additional example uses a different, explicitly stated set of assumptions. Do not combine its inputs with the main example above.
A profitable-looking ROAS can still lose money
Suppose the order value is $50 and non-ad costs total $35, leaving $15 contribution. Break-even ROAS is $50 ÷ $15 = 3.33×. A campaign with 3× ROAS spends $16.67 per $50 order and loses about $1.67 before any additional fixed overhead.
| Target | Maximum CPA | Required ROAS |
|---|---|---|
| Break even | $15 | 3.33× |
| Keep $5 profit per order | $10 | 5.00× |
| Keep $10 profit per order | $5 | 10.00× |
Translate CPA into a CPC limit
At an assumed 2% purchase conversion rate, a $10 target CPA supports a CPC of $0.20 because $10 × 0.02 = $0.20. If CPC is $0.30 and conversion rate stays at 2%, expected CPA is $15. The result changes when conversion quality, average order value or returns change.
Use consistent revenue definitions
Discounted revenue, refunded orders, attribution windows and tax/shipping treatment can make advertising-platform revenue differ from bookkeeping revenue. Use a consistent definition and deduct all modeled costs before interpreting ROAS as profit.
Load your actual costs into the Google Shopping profit calculator. Use Break-even ROAS for a simpler per-order model and Monthly Profit for fixed advertising and overhead. These examples are arithmetic scenarios, not a claim about campaign results or a Google bidding recommendation.