CPM vs CPC vs CPA: One Campaign, Three Prices, and the Two Rates That Connect Them
CPM, CPC, and CPA are the same $10,000 divided by three different denominators. Click-through rate and conversion rate are the bridges between them. Here is the chain worked end to end, and what a half-point CTR drop does to your cost per sale.
CPM vs CPC vs CPA
Three metrics, one budget. CPM is the price of a thousand impressions, CPC the price of a click, CPA the price of a customer. They are not alternatives; they are the same spend measured further and further down the funnel.
Two conversion rates connect them. Click-through rate turns impressions into clicks, and conversion rate turns clicks into customers. If you know the CPM and both rates, every other number in the chain follows arithmetically.
Try the daily ad budget calculatorWork backwards from a target number of conversions to the daily spend that delivers them.The chain, worked on $10,000
A campaign buys impressions at a $8.50 CPM. The creative earns a 0.90% click-through rate, and the landing page converts 2.40% of clicks. Average order value is $95 at a 60% gross margin.
Swipe sideways to compare columns.
| Step | Working | Result |
|---|---|---|
| Budget | — | $10,000 |
| Impressions | 10,000 ÷ 8.50 × 1,000 | 1,176,471 |
| Clicks | 1,176,471 × 0.90% | 10,588 |
| Effective CPC | 10,000 ÷ 10,588 | $0.94 |
| Conversions | 10,588 × 2.40% | 254 |
| Effective CPA | 10,000 ÷ 254 | $39.37 |
| Revenue | 254 × $95 | $24,130 |
| Gross profit | $24,130 × 60% | $14,478 |
| Profit after ad spend | $14,478 − $10,000 | $4,478 |
Why CTR matters more than the CPM you negotiated
The two rates sit in the denominator, so they move CPA harder than the media price does. Hold everything else and vary only the click-through rate.
Swipe sideways to compare columns.
| CTR | Clicks | CPC | Conversions | CPA | vs $57 break-even |
|---|---|---|---|---|---|
| 0.50% | 5,882 | $1.70 | 141 | $70.92 | Losing $13.92 each |
| 0.70% | 8,235 | $1.21 | 198 | $50.51 | Marginal |
| 0.90% | 10,588 | $0.94 | 254 | $39.37 | Working |
| 1.20% | 14,118 | $0.71 | 339 | $29.50 | Strong |
| 1.80% | 21,176 | $0.47 | 508 | $19.69 | Scale it |
Going from 0.50% to 0.90% CTR cuts CPA by 44%. Negotiating the CPM from $8.50 down to $7.00, an 18% media saving that would take real effort, cuts CPA by 18%. The creative is the bigger lever, and it is usually the cheaper one.
Conversion rate behaves the same way and compounds with it. At 0.90% CTR and a 3.20% conversion rate, CPA falls to $29.50. Improving both rates by a third each takes CPA to $22.15, a 44% cut, because the two multiply.
Try the lead conversion ROI calculatorModel how a change in conversion rate at any funnel stage flows through to cost per customer.Which pricing model to buy on
Swipe sideways to compare columns.
| You pay for | You carry the risk of | Platform carries the risk of | Suits |
|---|---|---|---|
| CPM, impressions | Bad creative, bad targeting, no clicks | Nothing downstream | Brand campaigns, proven creative, reach goals |
| CPC, clicks | A landing page that does not convert | Creative that nobody clicks | Most direct response, most of the time |
| CPA, actions | Very little | Everything up to the conversion | Affiliate deals, mature offers with clean tracking |
Risk transfer is priced. A CPA deal costs more per conversion than the same traffic bought on CPM would, because the seller is absorbing the uncertainty. That premium is worth paying when you cannot predict your own funnel and worth refusing once you can.
Reporting the chain, not the metric
A campaign report that says "CPA rose from $39 to $52" tells nobody what to do. The same report with the whole chain identifies the failure in one glance.
Swipe sideways to compare columns.
| What moved | Likely cause | What to change |
|---|---|---|
| CPM up, CTR flat | Auction competition or seasonality | Bid, budget pacing, or dayparting |
| CPM flat, CTR down | Creative fatigue or audience saturation | New creative, wider or fresher audience |
| CTR flat, conversion rate down | Landing page, offer, or checkout | Page speed, form length, price, stock |
| Conversions flat, revenue down | Discounting or product mix | Average order value, not the media at all |
| Everything flat, CPA up | Tracking loss | Check attribution before changing anything |
That last row deserves attention. A CPA increase with no change in any upstream metric usually means conversions are happening and not being recorded. Changing bids in response to a tracking problem makes a healthy campaign look worse.
Try the ROAS to ROI calculatorConvert a return on ad spend into an actual profit figure by putting cost of goods back in.And why none of the three is ROAS
The campaign above returned $24,130 on $10,000, a 2.41 ROAS. That sounds like a 141% return and is not. Cost of goods takes 40% of the revenue, so the real profit after ad spend is $4,478, a 44.8% return on the spend.
What these metrics do not tell you
- None of them accounts for repeat purchases. A $39 CPA against a $57 break-even looks thin and is excellent if a third of those customers buy again within a year. Compare CPA against lifetime value, not against a single order.
- Attribution decides the numbers. Last-click, first-click, and data-driven models can differ by 30% or more on the same campaign, and the CPA you report is a function of the model you chose.
- Averages hide the distribution. A blended $39 CPA can be one audience at $18 and another at $95, and the average tells you to keep funding both.
- Incrementality is not measured here. Some of those 254 conversions would have happened without the ad, particularly on branded search and retargeting. CPA does not distinguish caused from captured.
- Impressions are not attention. Viewability, fraud, and below-the-fold placement all mean a purchased impression may never have been seen, and CPM does not care.
- The funnel is treated as one step per stage. Real journeys involve several sessions and several devices, and the tidy chain above is a model rather than a description.
- Nothing here covers creative or offer quality, which drive the two rates that drive everything else.
How do I convert CPM to CPC?
Divide the CPM by 1,000 to get the cost per impression, then divide by the click-through rate. An $8.50 CPM at a 0.90% CTR is 0.0085 ÷ 0.009 = $0.94 per click. Halve the CTR and the CPC doubles.
What is a good CPA?
Anything below your break-even CPA, which is average order value times gross margin. There is no industry benchmark that means anything, because the same $40 CPA is excellent on a $200 product and fatal on a $30 one.
Should I buy on CPM or CPC?
CPC if you are unsure the creative will earn clicks, since the platform then carries that risk. CPM once you have a proven click-through rate, because it is usually cheaper per click at that point. Compare by converting the CPM offer to an implied CPC at your actual CTR.
Why did my CPA rise when my CPC fell?
Because cheaper clicks are often worse clicks. Broadening targeting or bidding down usually buys lower-intent traffic, so the conversion rate falls faster than the click price does. Watch CPA rather than CPC when you make that kind of change.
Is a 2.41 ROAS good?
It depends entirely on gross margin. At a 60% margin, break-even is 1.67, so 2.41 is profitable. At a 30% margin, break-even is 3.33, so the same 2.41 loses money on every sale. Always state the margin alongside a ROAS target.
Which metric should I optimise a campaign toward?
CPA, or better, profit per customer. CPM and CPC are diagnostics that explain why CPA moved. Optimising toward a cheap CPC reliably produces campaigns with excellent click prices and no sales.
Written by
Do The Calculation Team
Do The Calculation
Do The Calculation is built by a small team of data analysts and spreadsheet developers. Where a guide depends on a published formula, standard, or government rule, the calculator it links to names that source directly so you can check the number yourself.
About the team