CPM shows what an advertiser pays per 1,000 ad impressions. RPM shows what a publisher actually earns per 1,000 page views or impressions, after fill rate, ad units, and revenue share all take their cut.
According to Google’s official AdSense documentation, RPM “doesn’t represent how much you’ve actually earned” in a literal sense; it’s a calculated rate, just like CPM, but the two numbers answer different questions for different sides of the same transaction.
That distinction carries real weight in practice. Advertisers who confuse RPM vs CPM end up misjudging media cost. Publishers who track only CPM end up missing how much of that spend they actually capture. Understanding CPM and RPM side by side gives both parties a clearer read on what’s really happening in a campaign.
This article breaks down what each metric means, how to calculate them, and how publishers and advertisers should use them differently.
Get expert support and discover strategies for your campaigns perform better.
CPM vs RPM in One Table
The clearest way to see the difference between CPM and RPM is side by side, since the two metrics share a structure but measure opposite sides of the transaction.
| CPM | RPM | |
| Meaning | Cost per 1,000 ad impressions | Revenue per 1,000 page views or impressions |
| Formula | (Campaign cost ÷ Impressions) × 1,000 | (Revenue ÷ Page views or impressions) × 1,000 |
| Main user | Advertiser | Publisher |
| Measures | Cost | Revenue |
| Based on | Ad impressions delivered | Page views, sessions, or impressions (platform-dependent) |
| Best use case | Comparing media costs across inventory, formats, and audiences | Comparing monetization efficiency across pages, traffic sources, and layouts |
| Main limitation | Doesn’t reflect fill rate, viewability, or actual publisher earnings | Varies by platform definition and can mask which specific ad unit performed well |
What Is CPM and How Do Advertisers Use It?
CPM stands for cost per mille, the cost an advertiser pays for every 1,000 ad impressions delivered. The formula is straightforward:
CPM = (campaign cost ÷ impressions) × 1,000
Say a campaign spends $2,500 and delivers 500,000 impressions. The CPM would be ($2,500 ÷ 500,000) × 1,000, or $5.00. That single number lets an advertiser compare the cost of reaching an audience across very different inventory, a display network, a CTV placement, a programmatic deal, without needing to normalize for scale manually each time.
Advertisers lean on CPM to compare formats, audiences, and channels on equal footing. A video placement at $15 CPM and a display placement at $4 CPM tell a buyer immediately which one costs more to reach the same volume of impressions, even before factoring in performance. Publishers, on their side, watch CPM too, since rising CPM on their inventory signals stronger advertiser demand for that specific placement, audience, or content category.
A Short CPM Example
A finance brand runs a $10,000 display campaign that delivers 2 million impressions. CPM = ($10,000 ÷ 2,000,000) × 1,000 = $5.00. If a competing campaign on the same network delivers the same volume for $8,000, its CPM of $4.00 buys the same reach for less, all else being equal.
What Is RPM and How Do Publishers Use It?
RPM stands for revenue per mille, the estimated revenue a publisher earns per 1,000 page views, sessions, or impressions, depending on which version of the metric a platform reports. The formula mirrors CPM’s structure but flips the inputs:
RPM = (revenue ÷ page views or impressions) × 1,000
Google’s own AdSense documentation gives a clean worked example: earn an estimated $0.15 from 25 page views, and the page RPM equals ($0.15 ÷ 25) × 1,000, or $6.00. A separate ad-impression-based version works the same way: $180 earned from 45,000 ad impressions produces an ad RPM of ($180 ÷ 45,000) × 1,000, or $4.00.
RPM reflects what a publisher actually captures, not what advertisers were willing to pay before fees, fill rate, and revenue share got applied. A page can sit on premium inventory with strong advertiser CPMs and still post a mediocre RPM if the fill rate is low or several ad units are competing for the same limited attention on the page. That gap between advertiser cost and publisher capture is the entire reason RPM exists as a separate metric.
A Short RPM Example
A blog earns an estimated $20 in ad revenue from a page that received 2,000 page views in a month. RPM = ($20 ÷ 2,000) × 1,000 = $10.00. A second page on the same site earns $15 from 3,000 page views, for an RPM of $5.00. Even if both pages sit in the same ad network at similar CPMs, the first page is converting that advertiser demand into actual revenue more efficiently.
How Should Publishers Use CPM vs RPM Differently?
Publishers get the most value from CPM and RPM when they treat them as answers to two separate questions: what is my inventory worth, and what am I actually earning from it.
Why Publishers Should Track CPM
CPM tells publishers how much advertiser demand exists for their inventory at a given moment. A rising CPM on a specific placement or content category signals advertisers competing harder for that audience. That signal is useful in several concrete ways:
- Comparing ad formats, geos, devices, and placements against each other to see where advertiser demand concentrates.
- Pricing direct deals with a defensible number, since a publisher quoting CPM to a buyer needs to know where their inventory sits relative to the open market.
- Informing yield strategy by flagging which inventory segments are commanding premium rates worth protecting or expanding.
Why Publishers Should Track RPM
RPM tells publishers what they’re actually keeping after every factor between advertiser demand and final payout gets applied. That makes it the more useful number for revenue planning, since it accounts for fill rate, the number of ad units on a page, and platform revenue share all at once.
RPM is also the better lens for comparing performance across pages, content categories, and traffic sources, since two pages with identical CPM exposure can post very different RPM if one has a better fill rate or a layout that supports more viewable ad units.
Publishers working to raise yield without sacrificing user experience can find practical tactics in 10 tactics to increase ad revenue, several of which move RPM specifically rather than headline CPM.
The key publisher takeaway: CPM tells publishers what inventory may be worth, and RPM tells them what they actually capture. Tracking only one half of that pair leaves a real gap in how a publisher understands their own business.
How Should Advertisers Use CPM vs RPM Differently?
Advertisers generally live inside CPM day to day, but understanding RPM changes how some of those CPM decisions get made.
Why Advertisers Focus on CPM
CPM gives advertisers a direct way to estimate the cost of reaching a target audience at scale, which makes it central to planning and budget allocation across most ad formats.
- Estimating reach cost before a campaign launches, so the budget gets allocated to the channels delivering the most impressions per dollar.
- Running awareness-focused campaigns across display, video, CTV, and programmatic, where reach and frequency matter more than a direct response metric.
- Comparing audience quality and buying efficiency across vendors and platforms using one consistent unit.
Why Advertisers Should Still Understand RPM
RPM is worth understanding on the advertiser side, too, for a less obvious reason: it reveals how the publisher on the other side of a deal actually values their own inventory.
A page with a high RPM signals strong commercial intent from its existing advertiser base, but that same high RPM often comes with more competition for the placement, since other buyers have already discovered the same value. Understanding this helps advertisers in three specific scenarios:
- Direct deals, where knowing a publisher’s typical RPM gives a buyer a realistic sense of what the publisher needs to earn for the deal to make sense on their end.
- Private marketplace deals, where RPM context helps a buyer judge whether a quoted CPM floor reflects genuine inventory value or an inflated ask.
- Premium inventory negotiations, where a publisher’s RPM history on a specific placement is a stronger anchor for pricing than CPM alone.
Learn more about how Target CPM strategies can be calibrated once an advertiser has a realistic read on the inventory’s actual value to the publisher, not just its listed rate card.
CPM and RPM Difference: The Simple Version Worth Remembering
CPM is what advertisers pay. RPM is what publishers earn. Both metrics use the same per-1,000 structure, which is exactly why they get confused, but they sit on opposite sides of the same transaction and answer different questions.
The strongest ad strategy doesn’t pick one metric over the other. It looks at CPM and RPM together with fill rate, viewability, user experience, conversions, and revenue quality, since any one number in isolation can tell a misleading story about how a campaign or a page is actually performing.
See how CPM and yield data come together in one platform
Explore ASTRAD’s reporting tools built for publishers and advertisers managing both sides of the equation.
FAQ
Yes, and this happens more often than marketers expect. A high CPM reflects strong advertiser demand for an impression, but if the publisher's fill rate is low or only a fraction of ad requests actually get filled, the revenue captured per 1,000 page views can still come out modest.
Low viewability typically suppresses both metrics, since advertisers pay less for impressions they suspect users never actually saw, and publishers earn less when fewer of their served ads count toward billable, viewable impressions. Improving viewability tends to lift CPM and RPM together rather than trading one against the other.
RPM tends to matter more for revenue planning regardless of traffic size, since it reflects actual earnings rather than a theoretical rate. Smaller publishers especially benefit from RPM because it shows which pages or content categories are worth prioritizing without needing the scale to negotiate CPM directly with advertisers.
Knowing a publisher's typical RPM helps an advertiser judge whether a quoted CPM leaves the publisher a sustainable margin or is priced unrealistically low, which has a real bearing on the health of a long-term direct or PMP relationship. Bidding with that context tends to produce more stable inventory access than chasing the lowest CPM available in the moment.
Yes, combining the two gives a more complete forecast than either alone. CPM trends show where advertiser demand is heading, while RPM trends show how efficiently that demand is converting into actual revenue, and projecting both together produces a more grounded revenue estimate than extrapolating from CPM in isolation.
