CPM vs CPC: What You Are Actually Paying For

One auction, two invoices

Nothing separate happens inside the ad server when you choose CPC instead of CPM. There is one auction, it fills one slot, and the winner is the advertiser whose bid multiplied by the platform's estimate of user response is highest. The pricing model only decides which event the meter clicks on. Buy on CPM and the meter runs on impressions, so you carry the risk that nobody clicks. Buy on CPC and the meter runs on clicks, so the platform carries that risk — and prices it in. Neither model changes how many people see your ad, and neither makes an ad better or worse. They allocate the same uncertainty to different sides of the contract.

That is why the two prices are never independent of each other. A platform selling clicks has to guess how many impressions it will burn to produce one, and it charges accordingly; a platform selling impressions leaves that guess to you. The moment you write both prices on the same page and ask which is cheaper, you have asked a question that cannot be answered without a third number. That number is the click-through rate, and it is not a property of the pricing model — it is a property of your creative, your audience and the placement.

CTR is the exchange rate, and 1.6% is the crossover

Write the conversion once and keep it: CPC = CPM ÷ 1,000 ÷ CTR. A thousand impressions cost the CPM; the CTR tells you how many clicks those thousand impressions produce; divide and you have the price of one click. Run it backwards and it works too: CPM = CPC × CTR × 1,000. Every argument about which model is cheaper collapses into a single comparison once you have made both sides speak the same unit.

Set the two equal and the crossover falls out. An $8 CPM matches a $0.50 CPC when 8 ÷ 1,000 ÷ CTR = 0.50, which gives CTR = 8 ÷ 1,000 ÷ 0.50 = 0.016, or 1.6%. That is the whole decision rule, and it is worth stating in words: if you expect to beat 1.6% on this placement, buy impressions; if you expect to fall short, buy clicks. The sensitivity is brutal in both directions. At 3% CTR the impression buy costs $0.27 a click, almost half the click price. At 0.8% it costs $1.00, double it. The gap between a good ad and a mediocre one on the same placement is routinely that wide, which is why the same buyer can be right to choose CPM in one campaign and CPC in the next.

One campaign, costed both ways

Take $4,000 and a placement where your creative reliably does 2%. On the CPM side, $4,000 at an $8 CPM buys 500,000 impressions, because 4,000 ÷ 8 × 1,000 = 500,000. A 2% CTR turns those into 10,000 clicks, so your effective CPC is 4,000 ÷ 10,000 = $0.40. On the CPC side, $4,000 at $0.50 buys 8,000 clicks and nothing more is knowable — you do not control how many impressions it took. If it also ran at 2%, the platform served 400,000 impressions to deliver them, which is an effective CPM of $10. Same money, same placement, same creative: 10,000 clicks against 8,000, and an effective CPM of $8 against $10.

Push it one step further, because clicks are not the goal either. At a 3% landing-page conversion rate the impression buy produces 300 conversions and the click buy 240, so the cost per acquisition is 4,000 ÷ 300 = $13.33 against 4,000 ÷ 240 = $16.67 — a 25% penalty for having bought the wrong unit. The lesson is not that CPM wins. It is that the advantage propagates untouched all the way down the funnel: whatever multiple you gain in clicks, you keep in conversions, because the conversion rate applies equally to both piles.

Why the platform quotes one price and bills the other

Modern ad systems mostly bill on impressions no matter what you asked for. You set a target — a cost per click, a cost per acquisition, a return on ad spend — and the platform treats it as an objective, not a contract. It then buys impressions on your behalf, at whatever price the auction demands, and reports back an average cost per click that is an outcome rather than a rate you agreed to. This is why the CPC you see in the report drifts day to day even when your bid never changed, and why it can exceed the target you entered: the target constrains the average the optimiser aims at, not the price of any individual impression.

The practical consequence is that a quoted CPC is a forecast and a quoted CPM is a price. Treat them accordingly. When a media owner sells you a fixed CPM on a reserved placement, that is the number you will pay, and your risk is entirely in the CTR. When a self-serve platform shows you a CPC, it is describing what happened last week to advertisers who look like you, and your risk is that you do not look like them. Neither is dishonest, but they carry different kinds of certainty, and the honest planning move is to convert both to a cost per acquisition before comparing.

The viewability trap: two CPMs that are not comparable

An impression is counted when the ad is served. Whether a human could have seen it is a separate question, and the industry answers it with a separate metric: an impression counts as viewable when at least half its pixels are in the browser viewport for at least one continuous second for display, or two continuous seconds for video, per the Media Rating Council guidelines. Placements differ enormously on this. A unit high in an article that people actually scroll through can clear 80% viewability; a stack of units far below the fold on a page nobody scrolls may sit near half that.

Divide the price by the viewable share and the ranking often reverses. An $8 CPM at 80% viewability costs 8 ÷ 0.80 = $10 per thousand impressions a human could have seen. A $6 CPM at 50% viewability costs 6 ÷ 0.50 = $12. The cheaper CPM is the more expensive buy by 20%, and no amount of staring at the two rate cards reveals it. This is the single most common way a media plan goes wrong on paper before a single ad runs: two CPMs are compared as if they were the same product, when one of them is quietly selling you a larger fraction of impressions that never reached a screen.

A decision rule you can apply before spending

Start by putting both offers into the same unit. Convert every CPM to an effective CPC at the CTR you honestly expect, not the CTR from your best campaign of last year, and convert every CPC to an effective CPM so you can sanity-check it against what the inventory usually clears at. If the two conversions disagree wildly, one of your assumptions is wrong and it is almost always the CTR.

Then apply three rules of thumb that hold up in practice. Buy impressions when you are confident in the creative and the targeting, because you are effectively betting that your CTR will beat the crossover and you keep the entire upside. Buy clicks when you are testing something new, entering an unfamiliar placement, or working with a creative you have no history on, because you are paying the platform to absorb a risk you cannot yet price. And buy impressions, deliberately, when the campaign objective is not clicks at all — a brand campaign whose whole purpose is that people saw it should be measured and bought in the currency of viewable impressions, not dragged into a click auction that will optimise it toward whoever happens to click adverts.

Frequently asked questions

Is CPM or CPC cheaper?

Neither, until you supply a click-through rate. The two are linked by CPC = CPM ÷ 1,000 ÷ CTR, so the answer flips at a specific CTR that you can compute in one line. For an $8 CPM against a $0.50 CPC that crossover is 1.6%: 8 ÷ 1,000 ÷ 0.50 = 0.016. Above 1.6% the impression buy delivers cheaper clicks, below it the click buy does. Because the crossover moves whenever either price moves, there is no general rule that CPM is cheap and CPC is expensive, or the reverse. Compute the crossover for the two prices actually in front of you, then ask whether your creative and audience realistically beat it on that placement.

How do I convert a CPM into a cost per click?

Divide the CPM by 1,000 to get the price of one impression, then divide by the click-through rate expressed as a decimal. An $8 CPM at a 2% CTR gives 8 ÷ 1,000 = $0.008 per impression, then 0.008 ÷ 0.02 = $0.40 per click. The reverse conversion multiplies: a $0.50 CPC at a 2% CTR is 0.50 × 0.02 × 1,000 = a $10 effective CPM. Two cautions. Use the CTR of the specific placement, not a blended account average, because CTR varies more across placements than prices do. And remember that the effective CPC you compute this way is a forecast — it becomes a fact only after the campaign has run and you divide actual spend by actual clicks.

Why is my reported CPC different from the bid I set?

Because on most platforms the number you set is a target for an average, and the thing being auctioned underneath is still an impression. The system buys impressions at whatever the auction clears at, then divides your spend by the clicks that resulted and reports the quotient. That quotient moves with competition, with the mix of placements the optimiser chose that day, and above all with your realised click-through rate — a creative that fatigues will push your reported CPC up without anyone touching a bid. It is also why a reported CPC can land above the target you entered: nothing in the system promised you a per-click price, only an objective the optimiser was steering toward.

Can I compare two CPM quotes directly?

Only if both sellers count impressions the same way and deliver similar viewability, which they usually do not. An impression is recorded on serve; viewability asks whether at least half the ad's pixels were in the viewport for at least a continuous second, and placements vary widely on that. Divide each price by its viewable share before comparing. An $8 CPM at 80% viewable works out to $10 per thousand viewable impressions, while a $6 CPM at 50% viewable works out to $12 — the cheaper rate card is the more expensive buy by 20%. Ask for the viewability rate in writing, and if the seller cannot supply one, treat the quoted CPM as an upper bound on what you are getting rather than a price.

Does a low CPM mean the traffic is bad?

Not by itself, but a low CPM is always paid for with something. Impression prices fall when the audience is broad rather than targeted, when the placement sits below the fold, when the inventory is remnant rather than reserved, or when a large share of impressions is never viewable. Each of those depresses the CTR and the conversion rate as well, which is why a $2 CPM with a 0.1% CTR is a worse deal than an $8 CPM with a 2% CTR: the first costs $2.00 a click and the second $0.40. The correct test is never the CPM in isolation. Convert to an effective cost per click, then to a cost per acquisition, and compare there — that is the only level at which cheap inventory and good inventory can be ranked against each other.