Most questions about AdSense RPM start with a comparison: is mine good, and why is it lower than last month? Both assume RPM is a price. It is not. It is a ratio that sums up what happened, and the same site can show very different figures depending on the month, where its readers live and how it is measured. Once you treat it as an outcome, the drop that worried you usually has an ordinary explanation.
What AdSense RPM actually measures
Mediavine’s explainer “Why Does My RPM Change?” (28 February 2019) gives the formula plainly: RPM = Revenue / Sessions × 1000. It “gives you an idea of how much you are earning per 1000 sessions (or pageviews)” and, in Mediavine’s words, “isn’t actually the RATE that you earn”.
That last point is the one to hold on to. Advertisers bid for impressions and clicks; nobody bids for your RPM. The figure is calculated after the fact from what they paid and how much traffic you had, which is why it cannot be negotiated and why it responds to things that have nothing to do with ad prices.
Page RPM versus session RPM
AdSense reports page RPM: revenue per 1,000 pageviews. Mediavine prefers session RPM: revenue per 1,000 sessions. The revenue is the same in both; only the divisor changes. Because a session usually contains one or more pageviews, session RPM is at least as high as page RPM for the same site and period, and higher whenever readers view more than one page.
| Measure | Divides revenue by | Who reports it | What raises it |
|---|---|---|---|
| Page RPM | Pageviews, per 1,000 | AdSense | More revenue per page shown |
| Session RPM | Sessions, per 1,000 | Mediavine’s preferred measure | More revenue per visit, including extra pages per visit |
The practical consequence: never compare an AdSense page RPM with a figure someone quotes from a network that reports per session. A site averaging two pages per session shows a session RPM double its page RPM, with no change in what it earns. Many “network X pays twice as much” claims are this arithmetic and nothing else.
Why your AdSense RPM changes
Mediavine lists the reasons RPM moves. Grouped by whether you can do anything about them, they look like this:
| Driver | How it moves RPM | Yours to control? |
|---|---|---|
| Seasonal advertiser spending | Budgets rise and fall across the year | No |
| Monthly cycle | Dips in the first week, peaks at month end | No |
| Weaker months | January and July are named as softer | No |
| Traffic geography | Where readers are shapes demand; EU traffic needs a consent management platform | Partly |
| Dashboard settings | Ad density, unit frequency, category blocks | Yes |
| Engagement | Bounce rate and time on page | Yes |
| Site speed | Slow pages load fewer ads | Yes |
| Analytics problems | Broken or double-counted traffic distorts the divisor | Yes |
Read the first three rows before you change anything. If your RPM fell in the first week of January, the calendar is the likeliest cause, and redesigning your ad layout in response is solving the wrong problem.
The fix for the uncontrollable rows is how you compare, not what you change. Set your reports against the same period a year earlier rather than the previous month, so seasonal spending cancels out. Within a month, compare like with like: the first week against other first weeks, the last against other lasts. A month-on-month view will show a fall every January and every July whatever you do, and it will tempt you into changes that make the following month harder to read. If the year-on-year view is flat while the month-on-month view is falling, you have no RPM problem, only a calendar.
An RPM drop in the first week of January is usually the calendar, not your site.
Geography, consent and the “RPM by country” question
People search for AdSense RPM by country hoping for a table of dollar figures. We are not going to give one, because there is no verified source for it and any number would be out of date by the next season. What holds is the direction: advertiser demand differs by market, so a shift in where your readers come from changes RPM even if nothing else does.
Consent is the part of geography you can act on. Mediavine notes that EU traffic requires a consent management platform. If yours is missing or misconfigured, readers in those markets may see fewer or less valuable ads. Check your traffic-by-country report alongside revenue before deciding a drop is about ad prices.
How to increase AdSense RPM
Work through the controllable rows in order of risk, cheapest first.
- Fix measurement. Broken or double-counted analytics change the divisor. A tag firing twice makes RPM look worse than it is.
- Speed up the page. Ads that never load cannot earn. Test the templates that carry most of your traffic.
- Review dashboard settings. Ad density, unit frequency and blocked categories all trade revenue against reader experience. Change one at a time and give each change a full monthly cycle.
- Improve engagement. Pages that answer the query well and link onward keep readers longer, which gives ads more time on screen.
The received wisdom is to add more ad units. Sometimes that works; often it raises page RPM briefly while engagement falls, and session revenue goes nowhere. Judge every change on revenue per visit across a whole month, not on a single week.
RPM in context
RPM is one input to a site’s income, not the whole of it. Traffic times RPM is revenue, and traffic is usually the bigger lever. A site that doubles its readers at a flat RPM has done more than one that squeezes a modest RPM gain out of a denser layout, and it has not spent any reader goodwill doing it. Our ad network comparison covers when a move from AdSense makes sense, and Mediavine requirements sets out what that move needs.
To see how display-funded sites in a niche are monetised, open the database explorer, set monetization to display ads, filter by niche, and compare the ad networks the sites run against their estimated traffic.
The limitation is important here. No outside dataset can see a site’s RPM; estimated revenue figures are modelled, not reported, and should be read as a range. The traffic estimates post explains how far those models can drift. Your own AdSense reports are the only RPM figures you should act on.
If RPM in your subject is structurally low, the fix may be the model rather than the layout. The monetization hub and our guide to how websites make money cover affiliate and product income that does not depend on ad demand.
Frequently asked questions
What is page RPM in AdSense?
Page RPM is your estimated earnings per 1,000 pageviews. It is calculated from revenue and traffic after the fact, so it is an outcome, not a rate you are paid.
Why did my AdSense RPM drop?
Mediavine lists seasonal and monthly advertiser spending, weaker months such as January and July, traffic geography and consent, ad settings, engagement, site speed and analytics errors. Rule out the calendar and measurement before changing your layout.
Is session RPM higher than page RPM?
For the same site and period, yes, whenever readers view more than one page per visit. The revenue is identical; session RPM divides it by fewer units.
The takeaway AdSense RPM is revenue over traffic, not a price. Check the calendar, geography and your analytics before you touch ad settings, and compare page RPM only with page RPM.
