CPM vs CPC vs flat fee: which ad pricing model is right for you?

The CPM vs CPC vs flat fee question is really a question about risk. Every ad you can buy is priced one of three ways, and each model quietly assigns the risk of a bad outcome to someone different. Knowing which risk you are holding — and how to convert any quote into the other two models — is most of ad-buying literacy. (Disclosure: published by BidSurvivor, which is a flat-fee example below and ours, so weigh it accordingly.)

The three models, by who holds the risk

CPM (cost per thousand impressions): you pay for eyeballs, whether or not they act — you hold all the performance risk. Standard for display, video and most sponsorships. Cheap-looking numbers (a few dollars to a few tens of dollars per thousand) that can hide expensive outcomes, because a thousand impressions of a banner nobody looks at is a thousand of nothing. The Nielsen Norman Group has been documenting banner blindness since 1997; the CPM model is where that research sends your money.

CPC (cost per click): you pay only when someone acts, so the platform holds the showing risk and you hold the converting risk. Standard for search — the all-industry average is about $5.42 a click in 2026 (WordStream) — and much of social. The honest-looking model, with the caveat that platforms decide which clicks to find you, and the cheap ones are cheap for a reason.

Flat fee: one price for one placement for one period — a newsletter issue, a sponsorship month, a takeover day, an auction slot. All the risk is yours, and so is all the upside: a flat placement that overperforms costs the same as one that flops. The seller's incentive is to make you believe the audience number; yours is to check it.

The conversion maths (do it before every buy)

All three collapse into cost per visit. CPM ÷ (1,000 × CTR) is the cost of a click: a $10 CPM at 0.5% click-through is a $2 click. Flat fee ÷ expected visits is the same number from the other direction. Run every quote through this and the models compare directly. A "$50 flat" newsletter reaching 5,000 readers at 1% click-through is a $1 click, beating most CPC channels; the same $50 on a list that sends 10 visitors is a $5 click, beating nothing.

The number doing all the work in that sentence is the click-through rate, and it is the one the seller does not know either. Ask for last month's actual clicks, not the list size. If they will not say, assume the rate is a fraction of a per cent and price accordingly.

Worked example: $100 in each model

Take the same hundred dollars and buy it three ways, with round numbers so the arithmetic is visible.

• CPM. At a $10 CPM, $100 buys 10,000 impressions. Assume half a per cent click through — generous for display — and you have 50 visits at $2 each. Assume a tenth of a per cent, which is nearer to what an untargeted banner earns, and it is 10 visits at $10 each. Same invoice, five-fold difference in what you got.
• CPC. At the $5.42 average, $100 buys about 18 visits, each from a person who typed something related to what you sell. Fewer, warmer, and the platform's auction decides which eighteen.
• Flat. $100 buys one issue of a small newsletter, or two days on a niche site. Visits: whatever the audience does. Could be 5, could be 200. You find out afterwards, and only if the seller reports honestly.

The lesson is not that one model wins. It is that CPM is a bet on click-through, CPC is a bet on conversion, and flat fee is a bet on the seller's honesty about reach — and you should only take the bet you have data for. The mechanics of how the CPC and CPM prices are set in the first place are in how ad auctions set prices.

The costs no model quotes

Each model has a leak the rate card leaves out.

CPM leaks to fraud and viewability. Juniper Research expects roughly $100 billion of advertising spend to be lost to fraud in 2026 (Juniper); a bought impression that was never on a human screen still bills at full CPM.

CPC leaks to the platform's choices. A cost-per-click buy pays for clicks, but the platform picks the placements, the matches and increasingly the whole strategy — the September migration of Search campaigns to AI Max moved more of those choices out of the advertiser's hands, and whether you can still run a manual campaign is now a real question. You hold the converting risk on clicks you did not choose.

Flat fee leaks to the media kit. "Reaches 20,000 professionals" is a claim, not a measurement. Prefer sellers who publish their actual numbers, and read newsletter sponsorship vs display before signing a flat deal.

Where flat-fee auctions land

Our board is flat-fee with the price set by auction rather than by rate card. A BidSurvivor slot is twelve hours of the page — there are two a day — for whatever the auction clears: $0 for an empty slot, from $1.00 to take a held one, never above $5,000. Every visitor starts with $100 of house credit, no account needed, so early bids cost nothing out of pocket. The twist the other flat models lack is that every slot's card opens and click-throughs are published, on the front page and on each brand's page, so the "expected visits" in your conversion maths is a published number rather than a media kit's claim. Two rules to know before you rely on it: a brand that is outbid loses the slot and the money, with no refund, and a bid that fails to take the slot costs nothing. The scale is honestly small; the same slot priced per hour, per open, per click and per visit is worked through in what a twelve-hour ad slot costs.

Choosing in one paragraph

Buy CPC when you trust your landing page more than your targeting. Buy CPM only with the conversion maths done and a click-through rate you have measured, not hoped. Buy flat fee when the placement's audience is provably yours, and prefer flat sellers who show you last month's numbers — because "expected visits" is the whole gamble, and it should not have to be one.

If it helps, the ad budget calculator converts between all three at whatever budget you have. It is free and needs no account.