The formula, in one line
Breakeven ROAS = order value ÷ (order value − product cost − shipping − fees). The bottom half is your contribution margin: what one order leaves behind before you pay for the click. You break even when ad spend equals that margin, so revenue divided by margin is the ROAS that gets you to zero.
Every other ROAS target you have read is a guess about your costs. This one is arithmetic on your own numbers.
Why revenue-based ROAS misleads
A 4.0 ROAS sounds healthy. On a $60 order it means you paid $15 to get the sale and kept $45 of revenue. If that order cost $22 in product, $6 in shipping and $2.10 in fees, $30.10 of the $45 is already gone and you have $14.90 left, which is a $0.10 loss after the $15 ad spend. The dashboard called it a 4.0. Your bank account called it a loss.
This is the single most common reason a store scales ad spend and sees revenue grow while cash shrinks. The fix is not a better ad account, it is bidding against margin instead of revenue.
A worked example
| Line | Amount | Where it comes from |
|---|---|---|
| Average order value | $85.00 | Shopify, Online Store channel, last 30 days |
| Product cost | −$38.00 | Landed cost of the units in an average order |
| Shipping and fulfilment | −$7.00 | Label, packaging, pick and pack, return postage |
| Fees | −$3.40 | 4 percent of order value |
| Contribution margin | $36.60 | What an order leaves before ads |
| Ad spend at breakeven | $36.60 | Contribution margin per order |
| Breakeven ROAS | 2.32 | $85 ÷ $36.60 |
| Target ROAS at 12 percent profit | 3.22 | $85 ÷ ($36.60 − $10.20) |
That table is a different shop from the one loaded in the calculator, which is the point: change one input and the target moves. A store paying 8 percent in fees instead of 3.5 percent needs a higher ROAS for the same profit, and a store that has already built shipping into the price needs a lower one.
Two numbers to bid against
Max cost per purchase is your contribution margin. Pay that and you break exactly even. It is the ceiling, not a target: bidding at the ceiling means every bad month is a real loss.
Target cost per purchase is the margin after the profit you want to keep. This is the number to put into a target CPA bid strategy, and the number to compare against the cost per purchase in Ads Manager or Meta. If your actual CPA sits between the two, you are running a break-even ad account: busy, and not making money.
Where to get each number
| Number | Where to find it |
|---|---|
| Shopify orders, Online Store channel only | Orders → filter by date and channel. Include only orders that came through the website; POS and draft orders distort the average. |
| Product cost per order | Your supplier or landed cost for the units in an average order, not the retail price. |
| Shipping and fulfilment per order | Label cost, packaging, pick and pack, and returns postage. If a 3PL charges per order, this is that number. |
| Payment and platform fees | Card processing plus the platform's transaction fee, as a percentage of order value. Shopify Payments plus Shopify's fee is the usual 3 to 4 percent. |
| Profit you want to keep | A percentage of order value. Below 10 percent there is little room to absorb a bad ad month. |
The catch nobody mentions
All of this assumes the revenue in your ad platform is the revenue in your bank account. It often is not. A duplicated purchase event inflates both ROAS and the reported purchase value, so a store bidding against a doubled number quietly overpays for every order. The same happens when purchase events fire with an empty value, or when GA4 counts a purchase that Shopify later refunded.
Before you trust any ROAS target, reconcile: GA4 purchases against Shopify orders for the same 30 days, same time zone, Online Store channel only. Five to ten percent lower in GA4 is normal. Above 20 percent is a configuration error, and roughly double means duplicates. I wrote the full diagnosis, with the checks in order, in why GA4 shows Shopify purchases with a zero value, and the wider setup in the GA4 Shopify guide.
What this ignores
Returns and refunds, fixed costs like apps and subscriptions, the share of orders that never convert, and repeat purchases. A store with a 40 percent repeat rate can pay more for the first order than this calculator says, because the second order costs nothing to acquire. Add refunds as extra cost per order if your return rate is above a few percent, and treat the result as a floor rather than a promise.
Check the whole account, not one number
ROAS is the last number in a chain. If the purchase event is duplicated, if consent blocks half the traffic, or if the pixel and the Conversions API are both firing without deduplication, the margin you typed above describes a shop that does not exist. The fourteen checks I run on a client account are in the ecommerce tracking checklist, and the order of work is in conversion tracking for ecommerce.