Your ad report looks better. Revenue is up, ROAS has improved, and more orders are coming in. But after paying for the products, shipping, and marketing, you have less money left.
That can happen even when the reporting is accurate.
ROAS measures revenue attributed to advertising divided by ad spend. It doesn't tell you what it cost to fulfill those orders or run the business. A higher ROAS can come with a worse profit margin, and the extra revenue may not make up for it.
This came up in a recent Intensify client discussion. The business wanted to understand what advertising performance would support its profit goal after management fees and other costs. Looking at pricing changed the conversation. The question became whether the offer left enough money to acquire a customer and still earn a profit.
That is the question a ROAS target should help you answer.
Imagine an ecommerce business comparing two offers. One sells at its regular price. The other uses a larger discounted bundle that brings in more revenue but costs more to supply and ship.
The figures below are hypothetical. They illustrate the calculation, rather than report a client's results. Assume the revenue is confirmed, net of discounts and refunds, excludes sales tax, and is counted once.
| For the same period | Regular-price offer | Discounted bundle |
|---|---|---|
| Revenue attributed to ads | $30,000 | $40,000 |
| Ad spend | $10,000 | $10,000 |
| ROAS | 3.0 | 4.0 |
| Product, fulfillment, shipping, and payment costs | $12,000 | $28,000 |
| Contribution after ad spend | $8,000 | $2,000 |
The bundle has the higher ROAS and produces $10,000 more revenue. It also leaves $6,000 less to cover the remaining business expenses.
If those expenses are $6,000 for the period, the regular-price offer leaves $2,000. The bundle loses $4,000. For this simplified example, those are the only sales and costs in the business, before interest and income tax.
Choosing the bundle because it reached a 4.0 ROAS would reward the less profitable offer.
The calculation starts with the money you keep from each order before advertising. Use the price the customer actually pays, then subtract product costs and the variable costs of completing the sale. Account for refunds consistently so you don't subtract them twice.
Include the cost of a free gift and any shipping you cover. A bigger order can have a higher AOV while leaving less contribution if the discount and extra costs consume the increase.
Shopify's gross profit by product report is a useful starting point, but it subtracts recorded product costs from net sales. You still need to account for expenses outside that calculation. Shopify also notes that the report depends on having product costs recorded at the time of sale. Shopify's explanation of profit reports.
Once you know your contribution margin before advertising, you can calculate the ROAS needed to cover ad spend.
If you keep 50% of revenue before advertising, a 2.0 ROAS uses all of that contribution to pay for the ads. If you keep 25%, you need a 4.0 ROAS to reach the same point.
Neither calculation includes fixed expenses or a profit target. Calling either one “breakeven” without explaining what it covers can make an unprofitable business look healthy.
For a complete view, include management fees, creative production, software, payroll, and the other operating expenses relevant to the period. Some costs change as sales grow. Others stay fixed for a while. Your model should reflect how you actually pay them.
Then work backward from the profit you want at a realistic sales volume. A target that only works at a ROAS your business has never sustained may tell you that the offer needs to change.
A price increase could improve the contribution from each order. It could also lower conversion rate. A bundle might spread fulfillment costs across more products, or its discount might give away too much margin. Test what happens to contribution per visitor and total contribution as well as conversion rate and AOV.
Better creative can make more people want the product and understand its value. A stronger product page can answer the questions that stop them from buying. But every improvement still has to produce orders the business can afford to fulfill.
There is another reason a high ROAS can mislead: the people buying may already have been close to purchasing.
An ad reminding an existing customer about an offer has less persuading to do than an ad introducing the product to someone new. Branded search reaches people already looking for the brand. These sales can receive advertising credit even when the customer would have purchased without that particular ad.
Attribution assigns credit under a set of rules. Measuring additional sales caused by advertising requires a different question and, where practical, a controlled test. Google's Conversion Lift documentation explains how it compares people exposed to ads with a control group to estimate that effect. Google's explanation of attributed and incremental conversions.
This is why we look at what the advertising is doing for the business. Ads that explain how the product solves a problem can bring new people into consideration. Those people may later buy through a brand search or a reminder ad. Ranking every ad by reported ROAS alone can undervalue the work that introduced the customer.
Check total business revenue and contribution alongside platform reporting. Keep new-customer sales separate from repeat purchases. A blended result can look strong while acquisition weakens.
Repeat purchases can make acquiring a customer more profitable over time. They belong in the model when you have evidence for how often customers return, what they buy, and what it costs to serve them.
Use customers acquired in the same period and track their subsequent orders. Include the costs of those orders and any further marketing used to bring them back. Keep expected repeat purchases separate from purchases that have happened.
Timing matters too. A customer who becomes profitable after several months still requires you to pay for inventory and advertising before that profit arrives.
At your next performance review, ask to see how much contribution the orders produced after ad spend, what operating expenses remain, and how new customers are performing separately from returning customers.
If the answer is only a ROAS number, the profitability question is still unanswered.
