The screenshot usually arrives on a Monday. Return on ad spend: 3.0. Three dollars back for every dollar spent, sitting in green in Meta's Ads Manager or Google's campaign table, and it feels like the ads are working.
Maybe they are. But that number is revenue. It counts the whole price of the order, and the whole price was never yours to keep. If your store keeps 30 cents of every sales dollar after the product, the shipping, the box and the card fees, then three dollars of sales leaves you 90 cents. You paid a dollar for the ad. Every dollar you spent lost a dime, and the dashboard showed it in green.
Nothing is wrong with the ad account here. It is reporting exactly what it was built to report. What is missing is the line those numbers should be compared against, and every store has to work that line out for itself, because it comes from your margin, and the platforms never see your margin.
A Good ROAS Is a Number Only Your Margin Can Give You
Search for "what is a good ROAS" and you will find a lot of confident answers. 4x. 3x is the industry average. Anything over 2x is fine. None of them can be right for you, because none of them know what your products cost.
The number that matters is your break-even ROAS: the return at which the ad pays for itself and nothing more. Below it, every sale the ad produces loses money. Above it, every sale adds some. The math is one division.
Break-even ROAS = 1 ÷ your margin before ad spend.
If you keep 50 cents of every sales dollar, you need two dollars of sales to cover one dollar of ads, so you break even at 2x. Here is how quickly that line climbs as the margin thins:
- 60% margin: break-even at 1.67x
- 50% margin: break-even at 2x
- 40% margin: break-even at 2.5x
- 30% margin: break-even at 3.3x
- 20% margin: break-even at 5x
Two stores can post the same 3x in the same week. One is making real money. The other is paying for the privilege of being busy. Nothing in either ad account tells you which one you are.
One translation you will need. Google writes the same number as a percentage, so a 3.3x break-even shows up as 333% in a Google Ads target field. Google's own help page uses the example of $5 in sales for every $1 spent, which it writes as a 500% target. Meta writes it as a decimal: in its ROAS goal setting, a 110% return is entered as 1.100. Three notations, one idea. Keep your own line in "x" and convert it when you type it in.
Most Break-Even Calculators Leave Out Shipping, Card Fees and Returns. Your Real Line Is Higher.
The most popular way to work this out on YouTube goes like this. Take the sale price, subtract what the product cost you, divide by the sale price, and divide one by the result. A $60 order on a $22 product is a 63% margin, so you break even at about 1.6x. One widely watched version tells you plainly to use the cost of goods and nothing else.
That shortcut is where many owners get their number, and it is wrong in the direction that hurts. Every cost it skips makes the line look lower than it is, so campaigns that are quietly losing money get waved through as profitable.
Run the same $60 order with the costs that actually leave your account when it ships:
- Product: $22
- Shipping you paid and did not charge back: $8
- Mailer, box and insert: $1.50
- Card processing at 2.9% plus 30 cents: $2.04
- Returns, averaged across all orders: $2.50
- Discount codes, averaged across all orders: $2.40
That leaves $21.56 of every $60 order. A 36% margin, not 63%. The break-even line moves from about 1.6x to about 2.8x.
Now picture every campaign in the account sitting between those two numbers. A 2.2x campaign looks comfortably profitable against the shortcut. Against the real line, it loses around 21 cents on every dollar it spends, and it will keep doing so for as long as nobody redraws the line.
We have written a full walk-through of every cost that sits between gross profit and what you keep in You Know Your Best Seller. You Do Not Know Your Most Profitable Product., and the shipping piece alone gets its own treatment in Free Shipping Is Not Free. Use those to build the number. If most of your ad money goes to Google, our ecommerce guide builds the same line one cost at a time, with real return rates and shipping thresholds, in A 50% Margin Says You Break Even at 2.0 ROAS. Two rules apply specifically when the number is going to become a ROAS line.
Leave the ad spend out. When you calculate profit per product, ad spend belongs in the subtraction. When you calculate a break-even ROAS, it must not, because the ad spend is the thing the ROAS is paying for. Subtract it twice and your line comes out too high, and you will switch off campaigns that were making money.
Put your agency or freelancer's fee back in if it scales with spend. If someone manages your ads for 15% of what you spend, every dollar of ads actually costs you $1.15. Your break-even is then 1.15 divided by your margin, so the 36% store above needs about 3.2x, not 2.8x. A flat monthly fee works differently: it does not move the line per order, it has to be paid out of the profit above it.
The 3x in Your Ad Account Is Not the 3x in Your Bank Account
Even with the right line, you still have to compare it to the right number. The return in your ad account is built from a different set of facts than the money in your bank account, and the gaps between them almost always run in the platform's favor.
The sale value may include money that was never yours. Depending on how your store's tracking was installed, the purchase value sent to Google or Meta can include sales tax and the shipping fee the customer paid. Both pass straight through you. Check this once: pick a recent order, find it in your store, and compare its product subtotal to the conversion value the ad platform recorded for it. If the platform's number is bigger, your reported return is inflated by the same proportion.
Returns stay counted. The platform records the sale when the order is placed. If the customer sends it back three weeks later, the ad account still shows the sale unless someone sends the platform a correction. Google Ads accepts those adjustments. Very few small stores ever set them up.
Two platforms can claim the same order. A shopper sees your Instagram ad on Tuesday, searches your store name on Google on Thursday, and buys. Meta counts it. Google counts it. Your store recorded one order. Add the platforms' reported revenue together and you can end up with more sales than you actually made. We have covered why the counts disagree in GA4 Says 40 Conversions. Google Ads Says 61. Both Are Right., and what Meta's own sales measurement can and cannot tell a small account in Meta's Real Sales Test Needs 500 Conversions.
None of this makes the platform number useless. It is the number the bidding system steers by, so it is the number you set targets in. It just should not be the number that tells you whether you made money.
For that, there is a simpler check. Take your store's total sales for the month, net of returns and tax, and divide by everything you spent on ads across every platform. Compare that blended figure to your break-even line. It flatters the ads, because it includes repeat customers and people who would have found you anyway. That is exactly why it is useful: if even the flattering number sits below your break-even line, your advertising is losing money, whatever each dashboard says.
One ROAS Target Across Your Catalog Overpays for Some Products and Starves Others
Most stores do not have one margin. They have a range. The accessory you buy for $6 and sell for $35 sits next to the hero product whose supplier raised prices twice this year. And most ad accounts run all of it under one target.
Here is what that does. Take two products, both running at 3x:
- Product A has a 25% margin. Its break-even is 4x. At 3x, every dollar of ads brings in $3 of sales and 75 cents of margin. It loses a quarter on every dollar.
- Product B has a 50% margin. Its break-even is 2x. At 3x, every dollar brings in $1.50 of margin. It makes 50 cents on every dollar.
Same return, same report, opposite results. And a single target makes it worse in both directions. It is too loose for Product A, so the bidding system keeps buying sales that lose money. It is too strict for Product B, so the system holds back on sales that would have paid well.
Smarter Ecommerce, a bidding software company that works inside a large number of retail accounts, described this on its podcast in blunter terms than we would. They regularly audit accounts where an owner has carefully split campaigns by margin, ten of them, and then given all ten the same return target. Their verdict was "almost stupid." Under automated bidding, the target is the main lever you have over how hard each campaign bids. Building separate campaigns and then giving them the same target throws that lever away.
The fix does not need ten campaigns. For most small stores, three margin bands is plenty:
- Thin: under 30% margin. Highest target, or no ads at all if the break-even is out of reach.
- Middle: 30% to 45%.
- Rich: over 45%. Lowest target, and the band that should get the most room to spend.
On Google, the simplest way to get there is a custom label in your Merchant Center product feed that marks each product's band, then a campaign, or a product grouping inside one, for each band with its own target. We cover the wider structure decisions in Performance Max for Ecommerce Is Three Decisions. On Meta, the equivalent is a product set per band in your catalog, each in its own ad set with its own ROAS goal.
One caution before you split anything. A target needs data to work. Google's minimum for using a ROAS target on Search and Shopping is only 15 conversions in the past 30 days, counted across the account, but the bidding learns from far more than that, and every split leaves each campaign with fewer sales to learn from. If a band would only produce a handful of sales a month, merge it with its neighbor. Two bands with enough data beat three bands where the bidding system is guessing.
Break-Even ROAS Is the Floor. The Target You Give Google and Meta Sits Higher.
Break-even is not a goal. A campaign running exactly at break-even works hard, ships boxes, answers customer emails, and makes nothing. Rent, software and your own salary still have to come from somewhere.
So the target you actually type in starts from the break-even line and adds the profit you want each order to leave behind. The formula changes by one subtraction:
Target ROAS = 1 ÷ (your margin before ad spend − the profit you want to keep per sale).
Back to the store with the 36% margin. Suppose it wants each order to leave 10% of the sale as profit after the ad is paid for. That leaves 26 cents per sales dollar to spend on advertising, and 1 divided by 0.26 is 3.85. The target is about 3.85x, which goes into Google as 385% and into Meta's ROAS goal as 3.850.
Two things to know before you type that in.
First, a target is an instruction, not a hope. Google's bidding steers toward the number you set, and Meta's help center tells you to set a ROAS goal at "the lowest ROAS you can accept." Set it far above anything the campaign has ever done and it will not reach it. It will bid on less and spend less. Google itself says to set the target from your business goals with your recent results as a reference, and to move it gradually. We go further into how literally Google now reads that number in Your Google Ads Bid Target Isn't a Wish. It's an Instruction.
Second, set it against the platform's numbers, not your bank's. If you found in the last section that your platform's sale values run 10% high because tax and shipping are inside them, your target in the platform has to be 10% higher too, or you will be steering toward a return you never actually receive.
A 6x ROAS When 2.5x Already Pays the Bills Means You Are Turning Down Profitable Sales
Everything so far has been about campaigns that look better than they are. The opposite mistake is just as common in small stores, and it costs real money too.
If your break-even is 2.5x and a campaign is running at 6x, that is not only good news. It means the bidding system is buying only the cheapest, most certain sales it can find, and walking away from a large number of sales that would still have paid for themselves several times over. A very high return on a small budget usually means the budget, or the target, is holding the campaign back.
The agency behind a profit-reporting tool called Profit Compass describes exactly this with a client: a store holding a blended return of four to five times against a break-even they put near 2.5x, with spend capped at roughly the same amount every month. When they raised spend sharply for the holiday season, each new customer cost more, as they always do, and total profit grew anyway. That is their account of their own client, not an independent audit, but the arithmetic behind it is sound. Each extra dollar buys a slightly more expensive customer, and as long as that customer still lands above your line, the extra dollar was worth spending.
Google recommends the same move in its own guidance on targets: if you want more volume, lower the target gradually. The trap is reading the falling ratio as a problem. The average return across the whole budget will drop as you spend more. What matters is whether the last dollar you added still earns more than it costs, and we walk through how to check that in Google Ads Stops Growing Before Your Budget Does.
There is one more reason a store might deliberately run below its line on a first order: the customer who comes back. If your buyers reliably order again, a first sale that only breaks even can still be the best money you spend. That is a real strategy, with a real catch, which is that the profit arrives months after the ad bill does. We cover both halves in Your First Sale Barely Breaks Even. The Second One Is the Profit. and Lifetime Value Does Not Pay This Month's Invoice. For a small store without a long record of repeat orders, break-even on the first order is the safer floor.
Finding Your Break-Even ROAS Takes One Afternoon. Every Target You Set After That Rests on It.
You do not need software for this. You need your store's order export, your carrier invoices, and a spreadsheet.
- Pull 90 days of orders. Ninety rather than thirty, so one promotion does not set your line.
- Work out your margin before ad spend. Product cost, shipping you absorbed, packaging, card fees, returns and discounts, all as an average per order. Leave the ads out. Add a percentage-of-spend agency fee back in if you pay one.
- Divide 1 by that margin. That is your break-even line. Write it on a sticky note on your monitor.
- Check what the platform counts as a sale. Compare one real order to the conversion value it was recorded at, and note whether tax and shipping are inside it.
- Sort every campaign or product group into three piles. Above the line, between the shortcut line and the real line, and below. The middle pile is where the money has been leaking.
- Set targets from the line, not from habit. Break-even plus the profit you want to keep, by margin band where you have the volume to support it.
- Repeat it when your costs change. A supplier price rise, a new carrier contract or a new discount program moves the line, and the targets need to move with it.
This is also the first thing we work out with any online store before we change a single bid, because every decision about what to spend sits on top of it. If you would like a second set of eyes on the number, or on what your account looks like measured against it, that is the kind of work our ecommerce marketing team does every week.
Most owners have a number in their head for what a good return looks like. Very few of them worked it out. The afternoon it takes to replace that number with your own is the cheapest money-saving exercise in your whole ad account.
