How to Calculate Cost per Acquisition and Break-Even CPA

Cost per acquisition (CPA) is your ad spend divided by the number of orders it brought in. The most you can pay for one order without losing money is your break-even CPA: the profit an order leaves before advertising. Work it out as average order value × margin, then subtract what you pay per order for delivery, packaging, payment fees and returns. If your CPA is above that ceiling, the average order loses money, and if it is below, the gap is your profit per order.

CPA is one piece of the unit economics of an online store, next to average order value, margin and conversion rate. This guide covers the cost of one order: how to calculate it, why the figure in Google Ads can differ from yours, and where the ceiling sits.

How do you calculate cost per acquisition?

Divide what you spent on ads by the number of orders those ads brought in, over the same period:

CPA = ad spend ÷ orders

Google Ads reports this figure as cost per conversion. The Google Ads API reference (metrics) defines it as the cost of ad interactions divided by conversions. Shopify’s guide (Cost Per Acquisition: What It Is and How To Calculate) uses the same formula: total campaign cost divided by the number of conversions.

CPA, cost per order, cost per purchase and cost per conversion are the same number as long as a conversion means a paid order. Once the account counts something else as a conversion, they drift apart.

Why can the CPA in Google Ads differ from your cost per order?

The CPA in your account is only as accurate as the conversions behind it. Four things push it away from what an order really costs you.

1. The account counts something other than a purchase. According to the Google Ads API reference, the conversions metric includes only actions set to count in Conversions. Bid strategies optimise for those. If an add to cart or a page view sits there too, CPA falls, but it no longer means cost per order.
2. The account counts one purchase per click. Google offers two counting options for each action: every conversion, or one per ad click (About conversion counting options). Google calls “every” a good choice for sales, and it is the default for website actions. With “one”, repeat purchases after the same click drop out and CPA rises.
3. One order is counted twice. A second primary goal, such as an import from analytics or an old tag, fires on the same purchase. The account then shows more orders than there were, and CPA looks lower than it is.
4. The latest days are still filling in. Orders arrive after the click, so the most recent days show spend without all of their conversions. Google’s help page on Target CPA bidding says its recommended target is adjusted for these conversion delays.

You can check this in a few minutes. Take the same 30 days and compare purchases in Google Ads with the total orders your store received. A few orders can fall on either side of the period’s edges, because an order can come days after the click. If Google reports clearly more purchases than your store had orders, the account counts something besides purchases, or counts orders twice.

Sometimes CPA is the steadier number. If the order amounts reaching Google are wrong, revenue and ROAS are wrong too, but the number of orders can still be right. In that case the portal’s conversion check hides ROAS, shows revenue in grey marked “not trusted” and moves the focus to orders and cost per order.

What is break-even CPA, and which costs go into it?

Break-even CPA is the most you can pay for one order without losing money on it. It equals the profit one order leaves after the cost of goods and every other cost that comes with the order, before advertising. It is also called maximum CPA or allowable CPA.

Break-even CPA = average order value × margin − other costs per order

You can also write it as average order value × contribution margin, where contribution margin already has the other costs taken out.

This is how store owners put it to us. One said that to break even they would ideally need a CPA of UAH 150. Another described the loss: a product earned UAH 120 in profit, but one conversion cost UAH 200.

Which costs to subtract:

Shopify’s break-even guide (Break-Even ROAS Calculator: How To Measure Break-Even ROAS) lists the same per-order costs: cost of goods, delivery, payment processing and pick-and-pack fees. It leaves fixed costs such as rent and salaries out of the calculation.

Margin after the cost of goods alone is gross margin. Margin after every per-order cost is contribution margin, and that is the one that belongs in break-even CPA. The guide to contribution margin vs gross margin explains the difference and which costs sit between the two.

Shopify’s CPA guide suggests comparing CPA with customer lifetime value. That works when you know how often your buyers come back. Until you have that number, use the first order as the safe base.

A worked example

Example store, not client data.

A tableware store spends 40,000 a month on ads and gets 300 orders, so its CPA is 40,000 ÷ 300 = 133. Its average order value is 600 and its gross margin 35%, so each order leaves 210 after the cost of goods. Delivery and packing take another 5% of the order value, payment fees 2% and returns 3%: 60 per order in all.

The store is in profit by about 17 per order. If it used gross margin alone, its ceiling would be 210. A CPA of 200 would then look safe, but each such order would lose 50.

Break-even CPA and break-even ROAS are the same line

ROAS equals average order value divided by CPA, so a CPA ceiling is also a ROAS floor. For the example store, 600 ÷ 150 = 4.0, which is a break-even ROAS of 400%. Its actual ROAS is 180,000 ÷ 40,000 = 450%, above that floor. Use CPA when you think in orders and ROAS when you think in revenue: the answer is the same.

Why does every product have its own break-even CPA?

One store-wide figure averages orders that earn very different amounts. Order value and margin change from product to product, and break-even CPA changes with them.

Example store, not client data.

At the example store’s average CPA of 133, the cookware order leaves 267 and the mug order loses 80.5. The account average hides both.

This is why a cost-per-order target fits some stores better than others. A target CPA asks Google for one average cost per conversion, and every order counts the same towards it. That suits a store whose orders are close in size. When order values vary widely, read CPA next to order value or by product group.

A product has been tested once it has spent one order’s profit

Break-even CPA also tells you when to stop testing a product. In a published case on cutting wasted ad spend, an online store built its testing rule on the profit of one order. Its average order value was 1,150 Kč and its margin about 35%, so one order left 1,150 × 35% = 402.5 Kč in profit. A product counted as tested once its ad spend reached 400 Kč or more, and with no sale by then, the team marked it as loss-making.

The team noted that some owners allow 2–3× the margin for a test, and called that too much for their store. At 2–3×, a product that never sells has cost two or three orders’ worth of profit before anyone stops it.

The same logic works for any store. When a product has spent its break-even CPA without an order, it has already cost as much as a sale would have earned. Every further round of the same spend adds another order’s profit to the loss, so set the limit before the test starts.

Cheap products run out of room fast. At the example store’s cost per click of 5, a mug with a break-even CPA of 52.5 gets about ten clicks before it reaches its limit. The same case set a related rule: it kept add-on products out of its ads and promoted the main products they are bought with.

How do you bring CPA below break-even?

CPA has two parts: what a click costs and how many clicks turn into orders.

CPA = cost per click ÷ conversion rate

In the example store, 5 ÷ 3.75% = 133. Turn the formula around and you get the most you can pay per click: break-even CPA × conversion rate, or 150 × 3.75% = 5.63. That ceiling has its own guide: break-even CPC.

In our data, ROAS moved more with what happens after the click than with the auction. Within a store, month to month, ROAS went together with conversion rate (correlation +0.527) and with average order value (+0.540), and barely with cost per click (−0.100). That is GetProfit data on 1,360 store-months, June 2025 – June 2026, and it is an observation, not an experiment. ROAS is average order value divided by CPA, so these are the same levers seen from the revenue side.

Three ways to close the gap:

1. Raise the conversion rate. The product page, the price, delivery terms and stock decide whether a click becomes an order.
2. Raise the ceiling instead of lowering CPA. A higher average order value or lower per-order costs lift break-even CPA. Bundles and a free-delivery threshold are common ways to raise order value.
3. Move spend away from products that lose money on each order. Look for products or categories whose CPA sits above their own ceiling, such as a cheap item that earns less per order than the store’s average CPA.

How should a target CPA relate to break-even CPA?

Break-even CPA is the ceiling for your target CPA. The target sits between your current CPA and that ceiling, and the gap to the ceiling is the profit you keep per order.

Google suggests starting from history. According to Google Ads Help (About Target CPA bidding), a campaign with conversion data gets a recommended target CPA. It equals the average CPA of the last 30 days, adjusted for conversion delays.

Some conversions will cost more than the target and some less. Google aims to keep the average at the target. Google also warns that a target set too low may make you miss clicks that would have converted, so you get fewer conversions.

In the example store, actual CPA is 133 and the ceiling is 150. A target near 133 matches what the campaign delivers now and leaves about 17 per order. A target of 110, chosen to keep 40 per order, sits below what the campaign delivers today and risks fewer orders.

Two practical rules:

• Judge on enough data. Google recommends measuring performance over the last 30 days, with at least 30 conversions.
• Give the target room to work. The portal treats a daily budget of 3× the target CPA as the minimum for a campaign on that target. At a target of 133 that is about 400 a day, and the example store spends about 1,300 a day.

What to do this week

1. Check what counts as a conversion. Make sure only purchases feed the Conversions column, and that the account counts purchases as “every”.
2. Compare with your store. Over the same 30 days, Google’s purchases should be at or below your store’s total orders, give or take a few at the edges of the period.
3. Work out break-even CPA. Average order value × margin, minus delivery, packaging, payment fees and an allowance for returns.
4. Calculate actual CPA. Ad spend ÷ purchases over the last 30 days, leaving out the most recent days while conversions are still arriving.
5. Repeat by product group. Price bands or categories with different margins get their own ceilings. Find the groups whose CPA sits above theirs.
6. Set a test limit per product. One order’s profit is the rule from the published case. Decide in advance what happens to a product that spends it without a sale.
7. Set or check your target CPA. Start near your actual CPA for the last 30 days, keep it below break-even CPA, and give the campaign a daily budget of at least 3× the target.

See what share of your numbers you can trust. The portal checks whether the real order amount arrives together with the order, or every order carries one and the same number. The portal changes nothing without your consent.

Check your conversions →

Sources

• About Target CPA bidding — target CPA is the desired average cost per conversion; some conversions cost more and some less; the recommended target is the average CPA of the last 30 days, adjusted for conversion delays; a target set too low may bring fewer conversions; measure over 30 days with at least 30 conversions. Checked 2 October 2026.
• About conversion counting options — the “every” and “one” counting options; “every” suits sales and is the default for website actions. Checked 2 October 2026.
• metrics — Google Ads API reference (v25) — cost per conversion is the cost of ad interactions divided by conversions; only actions included in Conversions count, and bid strategies optimise for them; cost per all conversions uses all conversions. Checked 2 October 2026.
• Break-Even ROAS Calculator: How To Measure Break-Even ROAS — which per-order costs Shopify includes (cost of goods, shipping, payment processing, pick and pack) and that fixed costs stay out. Checked 2 October 2026.
• Cost Per Acquisition: What It Is and How To Calculate — the CPA formula; comparing CPA with customer lifetime value. Checked 2 October 2026.
• GetProfit data: 1,360 store-months, June 2025 – June 2026 — how ROAS moved with conversion rate, average order value and cost per click within a store.
• GetProfit published case — the testing rule of one order’s profit (1,150 Kč × 35% = 402.5 Kč; a product is tested at 400 Kč of ad spend) and the 2–3× alternative.
• GetProfit portal rules — a daily budget of at least 3× the target CPA.