Average CPM: What You Should Pay for a Thousand Impressions

A buyer watches CPM climb from $11 to $19 over three weeks and starts rewriting creative. Another sees $6 and calls the account healthy. Neither number means anything yet, because CPM is a price, and a price only says something next to what the thing is worth to you. A $19 CPM is cheap for a business earning $300 a customer and ruinous for one earning $12.

The published averages are still worth reading, so they are below. But the number that decides whether to keep spending is not an average. It is a multiplication you can do in about a minute with figures already in your account.

What the published averages actually measure

CPM is cost per thousand impressions, and it is the closest thing digital advertising has to a wholesale price. What it prices is not attention in general. It is attention from a particular audience, in a particular placement, on a particular week, against everyone else bidding for that same person. Change any of those and the price moves, which is why the ranges below are so wide.

Treat those as a map of neighbourhoods, not a scoreboard. They come from aggregated benchmark reports that blend thousands of accounts across industries, geographies and objectives, and the spread inside any one row is usually larger than the gap between two rows. A finance advertiser and a candle shop both running Meta feed will not sit anywhere near each other, and neither will sit on the average.

The one comparison that is safe is your own account against itself, week over week, holding the objective and the placement mix constant. Almost every CPM shift that gets blamed on the platform turns out to be a mix change.

CPM, CPC and CPA are one number in three costumes

Nothing about CPM is separable from the rest of the funnel, and the arithmetic connecting them is short enough to keep in your head.

Effective CPM = cost per click x click-through rate as a percent x 10

Break-even CPM = contribution per order x CTR x conversion rate x 1000

Take a store with an average order value of $86 and a contribution margin of 42 percent after goods, shipping, packaging, processing and a returns provision. Each order contributes $36.12. It runs Meta prospecting at a $14 CPM, a click-through rate of 1.1 percent and a site conversion rate of 2.5 percent.

A $14 CPM buys a thousand impressions, of which 11 become clicks, so each click costs $1.27. At a 2.5 percent conversion rate it takes 40 clicks to make an order, so the cost per acquisition is $50.91 against $36.12 of contribution. That account loses $14.79 on every order it produces, and it has been doing so from the first day, quietly, at whatever budget it was given. The CPM calculator and the CPC calculator will walk your own figures down the same chain.

The CPM you can afford is a multiplication

Run the chain backwards and you get a ceiling that belongs only to you. A thousand impressions at 1.1 percent CTR and 2.5 percent conversion produce 0.275 orders. At $36.12 of contribution each, those impressions are worth $9.93. That is the highest CPM this account can pay and break even, and it is paying $14.

Read the gap as work to be done, and notice that it can be done in three places. Cutting the CPM from $14 to $9.93 is a 29 percent reduction. Lifting the click-through rate from 1.1 to 1.55 percent is a 41 percent gain. Lifting the conversion rate from 2.5 to 3.52 percent is also 41 percent. All three are the same factor of 1.41 wearing different clothes, because CTR and conversion rate enter the formula in exactly the same way.

Which one you attack should be decided by which is cheapest to move, and the honest answer is usually the CPM. Shifting budget out of a narrow retargeting audience or switching a placement costs an afternoon. A 41 percent conversion rate lift is a quarter of work and often does not arrive. Before you commit to any of them, settle what your contribution per order really is in the contribution margin calculator, because an optimistic margin sets the affordable CPM too high by precisely the amount you flattered yourself.

What moves your CPM, roughly in order of size

Audience size is the biggest lever and the least discussed. A retargeting pool of 40,000 people has a hard ceiling on impressions, so as you spend into it the auction is mostly you against yourself, and frequency and CPM rise together. When a CPM doubles after someone tightened targeting to improve quality, nothing went wrong with the platform. The pool got small.

Seasonality comes next and is larger than most planning allows for. Meta and Google CPMs in the last three weeks of November routinely run 30 to 70 percent above October, because retail budgets arrive all at once against the same inventory. A campaign that breaks even in September can be a loss-maker in Q4 at identical creative and identical conversion rates, which is an argument for setting your affordable CPM before the season rather than during it.

Placement mix moves the blended figure without anything real changing. Reels, Stories and Audience Network clear well below the main feed, so a campaign that drifts toward cheap placements shows a falling CPM and often a falling conversion rate at the same time. That combination is not a win, and reading CPM alone will tell you it is.

Objective matters in the same direction. A conversion-optimised campaign deliberately buys expensive impressions, since it is asking the platform to find likely buyers rather than cheap eyeballs. Comparing its CPM against a reach campaign is comparing two different purchases. Creative quality sits last on this list, not because it does not matter, but because on auction platforms good creative shows up first as a higher click-through rate and only indirectly as a lower effective price.

The same budget, five prices

What the price does to a plan is easier to see in units than in percentages. Here is $10,000 spent at five different CPMs, held at the same 1.1 percent click-through rate and 2.5 percent conversion rate, with the resulting orders.

At $9 the budget returns $11,038 of contribution against $10,000 of spend, which is a thin profit. At $14 it returns $7,095, so the same money loses about $2,900. Nothing changed except the price of the impression, and no dashboard metric other than CPM would have shown you why. Plan the spend against a CPM you have decided you can pay rather than the one you got last month, which is what the ad budget calculator is for.

When paying far above average is correct

A high CPM is only a problem when the contribution behind it is small. LinkedIn at $70 is rational for a company with a $40,000 contract value and an 8 percent close rate on qualified leads, since one closed deal covers a great deal of expensive reach. Retargeting at $38 is rational when that pool clicks at 1.8 percent and converts at 6 percent, because a thousand of those impressions then produce 1.08 orders and carry $39 of contribution. Run the same multiplication on the audience you are actually buying and the ceiling moves with it.

The mistake is not paying a lot. It is paying a premium for an audience that behaves like a cheap one, which is what happens when a retargeting window is stretched to 180 days or a lookalike is built from a weak seed. In both cases the price stays premium while the behaviour reverts to prospecting. Check the click side of that story in the CTR calculator and the order side in the CPA calculator before you accept the premium.

Work out your ceiling before the next budget cycle

Take your last 30 days, pull the click-through rate and the site conversion rate for one campaign, multiply them by your contribution per order and by a thousand. That is the most you can pay for a thousand impressions on that campaign. Put it next to the CPM you actually paid, and you will know within a minute whether the account has a price problem, a funnel problem or no problem at all. Most buyers who do this for the first time find they have been trying to fix the wrong half.