How Conversion Rate Changes Your CAC: Worked Examples

Quick answer: For paid traffic, cost per customer is cost per click divided by conversion rate, so conversion rate sits underneath the fraction. In the illustrative example below (A$1.20 a click, A$60 left per order before ad cost), lifting conversion from 1.5% to 1.65% cuts cost per order from A$80.00 to A$72.73, a fall of 9.1%, yet the store still loses money on every order; it only breaks even at 2.0%. The number to watch is break-even headroom: what you can afford to pay for an order minus what you do pay. And a higher-converting audience is not automatically cheaper per customer, because its clicks can cost more and more of its buyers can be people who have bought before.

If your Meta or Google cost per customer keeps climbing, the first instinct is to look at the ads. That is fair. But there is a second input that does just as much work and sits on your own site: the share of people who click and then buy.

This post shows the arithmetic with fixed, made-up numbers, so you can see exactly what a small conversion lift does to your cost per customer (CAC) and to the room you have before an order loses money. Then it covers the caveat that catches people out: the audience or channel with the best conversion rate is not necessarily the cheapest way to get a customer. At the end there is a way to run the same sums on your own store in about half an hour.

A note on where we sit. Prism3 sells Shopify SEO, GEO and CRO. We do not run ads, email or SMS, so this post says nothing about how to build or bid campaigns. What it covers is the part of the sum that happens after the click, and the measurement that decides whether you can trust the numbers going in. Every dollar figure below is illustrative: round numbers chosen so the working is easy to follow. They are not benchmarks, not an average for any category and not a result from any store.

The formula: CAC = cost per click ÷ conversion rate

Start with how the platforms define their own numbers, because the formula falls straight out of them. These are the definitions as at 31 October 2026.

Now do the algebra. Over a month:

Spend  = clicks × cost per click
Orders = clicks × conversion rate
Spend ÷ Orders = cost per click ÷ conversion rate

The number of clicks cancels out. That is the whole idea: cost per order is the price of one click divided by the share of clicks that end in an order. At an A$1.20 click and a 1.5% conversion rate, that is A$1.20 ÷ 0.015 = A$80.00. Put differently, you buy 100 clicks for A$120 and get 1.5 orders from them.

It is the same thing Google’s two definitions give you. Cost per conversion is cost ÷ conversions, conversion rate is conversions ÷ clicks, so cost per conversion is cost per click ÷ conversion rate. Google’s page also notes that its cost-per-conversion calculation only covers eligible interactions, so clicks that cannot be tracked for conversions are removed from it.

When the formula equals your CAC, and when it does not

Shopify’s guide to customer acquisition cost defines CAC as all the costs of acquiring new customers in a period divided by the number of new customers who made purchases in that period, and it says that count is first-time customers only, not returning ones (shopify.com/blog/customer-acquisition-cost, read 31 October 2026). So the clean formula above equals CAC only if three things are true:

1. Clicks are your only acquisition cost. Shopify’s guide also lists marketing software, staff and contractor pay, content creation and discounts used to attract new customers.
2. Conversion rate means orders per click (more on why that is harder to measure than it sounds below).
3. Every order is from a first-time customer.

Where those fail, I call the result “cost per order” and treat cost per new customer as the stricter number. What counts as CAC on a Shopify store covers what belongs in the top and bottom of the fraction. Here I keep the inputs simple so the effect of conversion rate is easy to see.

Worked example 1: what a small lift does to CAC

The fixed inputs for every example in this post:

• Ad spend: A$9,600 in the month, on one paid channel.
• Cost per click: A$1.20, which buys 8,000 clicks.
• Starting conversion rate: 1.5%, which is 120 orders and a cost per order of A$80.00.
• Average order: A$110. After product cost, shipping, payment fees and packaging, A$60 is left per order before ad cost. I come back to that A$60 in the next example.
• Everyone is a first-time customer. I relax that later.

Now change only the conversion rate:

Three things worth noticing.

The effect is smaller than the lift, and the drop is bigger than the gain. Because conversion rate is the divisor, a relative lift of x cuts cost per order by x ÷ (1 + x). A 10% lift cuts it by 9.1%, a 20% lift by 16.7%, a 50% lift by 33.3%, and only a doubling halves it. Going the other way hurts more: a conversion rate that falls by a third pushes cost per order up by half. If your rate slips in a busy week, your cost per order moves further than you would guess.

Cost per click and conversion rate are mirror images. A 10% fall in cost per click cuts cost per order by exactly 10%: A$1.08 ÷ 0.015 = A$72.00. A 10% rise in conversion rate cuts it by 9.1%. So the two can cancel. If your conversion rate improves 10% to 1.65% in the same month your click price rises 10% to A$1.32, cost per order is A$1.32 ÷ 0.0165 = A$80.00, exactly where it started. Nothing in your ad account says the site did its job. Meta’s own page on cost per result lists the auction bid, target audience, optimisation type, ad creative and messaging, and schedule among the things that affect it. That list shows how many things sit behind the click price, which is why I track cost per click and conversion rate side by side, not just their ratio.

Budget does not appear in the formula. The spend cancelled out. That isolates the effect of conversion rate, but it says nothing about what happens to your click price if you spend twice as much. Your own account data is the only honest source for that.

Worked example 2: break-even headroom

Cost per order on its own tells you nothing until you set it against what an order is worth. The comparison is simple:

Break-even cost per order = profit per order before ad cost
Headroom per order        = break-even cost per order − actual cost per order

“Profit per order before ad cost” is the sale price minus everything that varies with the order: product cost, shipping you pay, payment fees, packaging, returns allowance. It is often called contribution margin per order. Contribution margin per order on Shopify, with a worked example shows how to build it from your own costs. In this post it is A$60 on an A$110 order.

If you will not lose money on a first order, your cost per order has to be at or below A$60. Rearranging the formula gives the conversion rate that gets you there:

Break-even conversion rate = cost per click ÷ profit per order before ad cost
                           = A$1.20 ÷ A$60 = 2.0%

The same sums as in example 1, now against A$60:

(“Month result after ad cost” is orders × A$60 minus the A$9,600 spend, so it ignores overheads and any later repeat orders. It is a first-order test.)

What the table says:

• The 10% lift is real progress and still a loss. Going from 1.5% to 1.65% closes about 36% of the A$20 per-order gap (A$7.27 of A$20), but the month is still A$1,680 under water. A report that says “cost per customer fell 9%” would be true and would hide that.
• The lift you need is a number you can compute. Required lift = break-even conversion rate ÷ current conversion rate − 1. Here that is 2.0% ÷ 1.5% − 1, about 33%. That is a much more useful thing to know than “improve conversion”.
• Close to break-even, small lifts matter more than they look. Headroom is the difference between two similar numbers (A$60 and A$80), so an A$7.27 move is a large share of it. The same lift on a store with an A$200 gap would barely register.
• The same CAC can be fine or a problem. A$72.73 per order is A$12.73 too high against A$60 of profit per order. Against A$90 of profit per order it would be comfortable. The verdict comes from the headroom, not the CAC on its own.

If you want the same logic expressed as a return on ad spend, break-even ROAS on Shopify walks through it. It is the same break-even in a different unit.

If some buyers are existing customers

Shopify’s CAC definition counts first-time customers only. Say 80% of the orders are first orders. Cost per new customer is then A$80.00 ÷ 0.8 = A$100.00 at a 1.5% conversion rate, and the break-even conversion rate for a first-order test rises to A$1.20 ÷ (A$60 × 0.8) = 2.5%. The more of your paid orders come from people who already know you, the further your real CAC sits from the platform’s cost per result.

A first-order test is the strict version

Plenty of stores accept a loss or a thin margin on the first order because they expect repeat purchases to pay it back. That is a legitimate choice, but it is a different calculation: it needs your repeat-purchase data and a view of how long you can wait for the cash. CAC payback period on Shopify covers the timing, and High CAC, Low LTV on Shopify: What to Do First covers how conversion rate, order value and repeat purchase each act on a different side of the problem. The sums in this post stay with the first order because it is the version nobody can argue with.

Worked example 3: a lift bought with a discount

A conversion lift is only as good as the margin it arrives with. Suppose the store runs 10% off to get the rate from 1.5% to 2.0%, the exact lift that broke even in example 2.

A 10% discount takes A$11 off the A$110 order. Assume product cost, shipping and packaging do not change (payment fees would fall slightly on a lower price, which I ignore), so profit per order before ad cost falls from A$60 to A$49. Everything about break-even moves:

At 2.0% with the discount, orders are 160 and cost per order is A$60.00 as before, but headroom is now A$49 − A$60 = −A$11 per order, and the month is 160 × A$49 − A$9,600 = −A$1,760. The same conversion rate that was break-even without the discount is a loss with it. To break even with 10% off, the rate has to reach about 2.45%, a lift of about 63% on the starting 1.5%.

This is not an argument against discounts. It is an argument for testing a lift against margin, not conversion rate alone. If a test shows a higher conversion rate, ask what happened to average order value and profit per order in the same period. Can improving conversion rate really double your Shopify revenue? covers the revenue side of the same trap, where a higher rate and a lower order value cancel each other out.

Worked example 4: a higher-converting audience is not automatically cheaper

This is the caveat that most worked examples skip. It is tempting to read “conversion rate is the divisor” as “the audience that converts best is the best audience”. It is not, because the other half of the fraction moves too.

Three invented audiences, same product, same A$60 profit per order (all figures illustrative):

Audience C converts about three times as often as A (3.5% against 1.2%) and still costs A$5 more per order. Audience B converts about twice as often as A and is the cheapest of the three. Neither “highest conversion rate” nor “lowest click price” picks the winner. Only the ratio does, and only compared against profit per order.

There are three reasons the best converter can lose.

1. Its clicks can cost more. A narrow audience can carry a higher click price than a broad one. Whether that is true for your audiences, your products and your market is a question for your ad account, not for this post.

2. More of its buyers are not new. An audience made up of people who already know your brand can convert well, and a share of those orders may be repeat customers. Shopify’s analytics fields reference lists “New customers” (first-time purchasers only) and “Returning customers” as separate counts. Adjust each audience for a returning share (invented here) and the picture changes:

Cost per new customer is cost per order ÷ the new-customer share, so C’s A$80.00 becomes A$80.00 ÷ 0.6 = A$133.33. Audience C may still be a good way to bring existing customers back. It is just not a cheap way to acquire new ones, and the two jobs should not share one number.

3. Cost per order ignores what the order is worth. If audience C’s buyers pick smaller baskets, their profit per order is lower than A$60, and the headroom shrinks further. Always compare to profit per order for that audience, not the store average.

The same reasoning applies to channels, not only audiences. A channel that reaches people already on their way to buy will show a high conversion rate. Whether it is “cheaper” depends on what it costs, how many of the buyers are new, and a question this arithmetic cannot answer: whether the order would have happened without the click.

Why your real numbers will not match the formula

The identity is exact on paper and approximate on a real store. Four reasons, each from the platforms’ own documentation (as at 31 October 2026).

A click is not a session. Meta’s help page on link clicks and landing page views says the two metrics differ because the page may not fully load, or the app may fail to open, after a link click, and because an ad may have more than one click destination. Cost per link click therefore understates what you pay for each visit that actually reached your store.

Conversion rate has three common definitions. Google Ads divides conversions by eligible interactions. GA4’s session key event rate divides sessions with a key event by sessions. Shopify divides sessions that completed checkout by sessions. Shopify’s Help Center notes that one customer can purchase more than once in a session, so your order count and your count of sessions that completed checkout may not match. See GA4 vs Shopify Analytics for how the two count.

Attribution windows differ. Meta’s “Results” page says its default attribution setting is 7-day click, 1-day view and 1-day engagement, and that in some cases results may be estimated or modelled. So Meta’s cost per result is spend divided by results counted that way, which is not the same as spend divided by Shopify orders. On Google’s side, the Ads help page says conversion columns are counted on the date of the click by default, offers “by conv. time” columns for comparing with tools such as Google Analytics or Shopify, and says discrepancies, often up to 20%, are expected because of different attribution models. Why Meta shows more sales than Shopify goes through the usual causes.

Shopify’s sessions have their own rules. Shopify’s fields reference says sessions are only counted when visitors consent to cookies through your cookie banner, and that conversion rate can look lower than expected because of bot traffic (it suggests filtering to human traffic). Separately, Shopify’s behaviour reports page says its Analytics session measurement rollout ran from 21 to 23 September 2026 and that, once the update reaches a store, sessions and conversion rate might look different without any change in traffic, orders or customer behaviour. If you compare conversion rate across that window, treat it as a new baseline, or you may credit (or blame) a real change for what was a measurement change.

The formula is still useful. Work out your numbers from one consistent source, and treat the platform’s own cost per result as a second opinion.

How to run this on your own store

About half an hour with a spreadsheet and three reports.

1. Pick one window and one source. A recent 28 or 30 days, one paid channel, and not a sale period if you can avoid it (keep sale weeks separate). Use the same dates for every number. A good Shopify conversion rate covers how to set a baseline.
2. Get spend from the ad platform for that channel and window.
3. Get sessions and orders for the same source from Shopify or GA4, using UTM-tagged links, not the platform’s click count. Note which attribution method the report uses.
4. Get new customers for the same source if your report can split them. If it cannot, say so, and use the platform’s figure with a caveat.
5. Calculate three numbers: cost per session (spend ÷ sessions), conversion rate (orders ÷ sessions), and cost per new customer (spend ÷ new customers).
6. Calculate profit per order before ad cost from your own costs.
7. Calculate break-even and required lift:

Break-even conversion rate = cost per session ÷ profit per order before ad cost
                             (divide by your new-customer share for a stricter test)
Required lift              = break-even conversion rate ÷ current conversion rate − 1

8. Compare your cost per order with the platform’s cost per result. If they differ a lot, that gap is itself a finding, and it is the first thing to understand before you change anything.

How to read the answer

The arithmetic gives you a required lift. What you do with it depends on how large it is and on what the rest of the store shows.

• If the required lift is modest, conversion work is a reasonable place to look first, because the fix sits on your own site and does not depend on an auction. A modest lift is still not a promise that it is available. Only looking at where visitors drop out tells you that.
• If the required lift is very large (a doubling or more), conversion rate alone is unlikely to close the gap, and the other inputs deserve attention: click price, profit per order (order value, bundles, discount depth), the new-customer share and the payback you can accept.
• If your numbers do not agree with each other, fix measurement before anything else. A required lift calculated on a wrong conversion rate is a wrong target. The CRO audit we run starts with quantitative analysis, including a tracking check, for exactly this reason.
• Do not treat a calculated lift as a forecast. I cannot tell you from this arithmetic what lift your store can get, and nobody can promise one.

Key takeaways

• For paid traffic, cost per order = cost per click ÷ conversion rate. It equals CAC only if clicks are your only cost and every buyer is new.
• A relative lift of x in conversion rate cuts cost per order by x ÷ (1 + x): 10% gives 9.1%, 50% gives 33.3%, and only a doubling halves it. A fall in conversion rate hurts more than the same rise helps.
• Cost per click and conversion rate offset each other. A conversion gain can be invisible if click prices rose in the same month.
• The decision number is headroom: profit per order before ad cost minus cost per order. In the illustrative example, break-even sat at 2.0% against a 1.5% start, a lift of about 33%.
• A lift bought with a discount moves break-even upward. Check profit per order alongside conversion rate.
• A higher-converting audience or channel is not automatically cheaper per customer. Check its click price, its share of new customers and its profit per order.
• Your real numbers will differ from the formula because clicks are not sessions, conversion rate has three definitions, attribution windows differ and Shopify rolled out a session measurement update in September 2026. Work from one consistent source.
• None of this predicts the lift your store can get.

FAQs

How does conversion rate affect customer acquisition cost?

For paid traffic, cost per order is cost per click divided by conversion rate, so a higher conversion rate lowers it and a lower one raises it. In the illustrative example here, a 10% relative lift in conversion rate (1.5% to 1.65%) cut cost per order by 9.1%, and a doubling halved it. The effect assumes click prices and the share of new customers stay the same.

What is the formula for CAC using cost per click and conversion rate?

Cost per order = cost per click ÷ conversion rate. It follows from Google Ads’ definitions (cost per conversion is cost ÷ conversions, and conversion rate is conversions ÷ clicks) and Meta’s (cost per link click is amount spent ÷ link clicks). It equals CAC only when clicks are your only acquisition cost and every order is from a first-time customer. Shopify’s guide defines CAC on first-time customers only.

If I double my conversion rate, does my CAC halve?

Yes, if cost per click, the share of new customers and everything else stay the same, because conversion rate is the divisor. In practice they rarely stay the same, and a doubling is a large step. A 10% lift cuts cost per order by about 9%, not 10%.

What is a break-even conversion rate and how do I work it out?

It is the conversion rate at which cost per order equals profit per order before ad cost. Divide your cost per click (or, better, your cost per session) by that profit per order. In the example here, A$1.20 ÷ A$60 = 2.0%. To see the lift you need, divide it by your current rate and subtract one.

Is a higher-converting audience or channel always cheaper per customer?

No. Cost per customer depends on the click price as well as the conversion rate, and on how many buyers are new. In the example, an audience converting at 3.5% cost A$80.00 per order against A$60.00 for one converting at 2.5%, and with 40% returning buyers its cost per new customer rose to A$133.33. Compare audiences on cost per new customer against profit per order.

Why does my ad platform show a different CAC from my Shopify numbers?

They count differently. Meta’s default attribution is 7-day click, 1-day view and 1-day engagement, and results can be modelled. Google counts conversions on the click date by default and says discrepancies, often up to 20%, are expected against tools such as Google Analytics or Shopify. Shopify counts sessions only with cookie consent and measures conversion as sessions that completed checkout over sessions. Pick one source for your CAC and use the platform figure as a cross-check.

Want a second opinion on where your store loses buyers?

If you would like a second opinion, book a free 30-minute diagnostic or tell us about your store. We share your Shopify store on screen and name two or three conversion leaks live, and tell you whether they are worth fixing. No pitch. If it makes sense to go further, our Shopify CRO service starts with a six-phase audit that includes a tracking check, so the numbers behind a calculation like this one can be trusted before anything is changed. We do not run ads or email.