Why reported ad RPMs vary so much between blogs
Search for what any ad network pays and you will find figures that contradict each other by an order of magnitude, all reported honestly by people describing their own dashboards. The variance is real, and most of it comes from four things that have nothing to do with the network.
1 August 2026 · 8 min read
RPM — revenue per thousand pageviews — is the number publishers compare, and it is close to useless as a comparison unless you know a great deal about both sites. That is not a criticism of anyone reporting theirs. It is a property of what the number is measuring.
Four variables account for most of the spread. We can quantify two of them from our own index and describe the other two precisely.
1. The publishers on a network are not alike
This is the variable people most often miss, and it is measurable. A network’s publisher base has a spread, and on some networks that spread is enormous.
| Network | Sites | Floor (p10) | Median | Top (p90) | Spread |
|---|---|---|---|---|---|
| Primis | 192 | 13 | 60 | 83 | 70 |
| Freestar | 637 | 16 | 53 | 78 | 62 |
| Snigel | 282 | 6 | 40 | 73 | 67 |
| Vidazoo | 443 | 1 | 39 | 79 | 77 |
| Publift | 2,090 | 5 | 36 | 70 | 65 |
| AdPushup | 997 | 1 | 33 | 74 | 74 |
| Raptive | 6,979 | 7 | 33 | 58 | 51 |
| Infolinks | 882 | 4 | 32 | 74 | 70 |
| Playwire | 968 | 3 | 31 | 71 | 68 |
| Adnimation | 564 | 3 | 30 | 70 | 67 |
| Mediavine | 8,638 | 10 | 29 | 47 | 38 |
| Setupad | 876 | 3 | 27 | 66 | 62 |
| Gourmet Ads | 972 | 3 | 22 | 70 | 67 |
| PubGalaxy | 2,599 | 1 | 18 | 68 | 67 |
| SHE Media | 1,591 | 4 | 17 | 40 | 36 |
| Venatus | 627 | 1 | 17 | 57 | 56 |
| Newor Media | 173 | 1 | 17 | 52 | 51 |
| NitroPay | 759 | 0 | 15 | 54 | 54 |
| Monumetric | 668 | 1 | 14 | 40 | 39 |
| Ezoic | 7,683 | 1 | 7 | 32 | 31 |
The spread column is the one to read. Vidazoo spans 77 points between its floor and its top decile; Ezoic, among the larger networks, spans 31. Reported figures from the first network will scatter far more than from the second, and that difference is a fact about the roster rather than about the rates.
2. Category matters more than most people expect
Advertisers pay for proximity to a purchase, and categories differ enormously in how close their readers are to one. A reader comparing appliances is worth a multiple of a reader looking up a game walkthrough, and no network can change that.
Our index shows the same divergence in the sites themselves: News & Politics sites have a median authority of 42 while Other sits at 12. Different subjects produce structurally different publishers as well as structurally different advertiser demand, and the two compound. The full breakdown is in the ad network landscape by niche.
3. Geography, which almost nobody states
Advertiser spending is concentrated in a small number of markets. An audience drawn mostly from those markets is worth several times one distributed across the world, at identical engagement.
This is the single largest unstated variable in most reported figures. Two food blogs with the same traffic and the same network can differ severalfold on this alone, and neither publisher will usually mention it because both regard their traffic as simply “their traffic”. If you are comparing yourself against a published number, the first question to ask is where that site’s readers are.
4. The month
Display rates are strongly seasonal. Fourth-quarter demand is materially higher than the spring trough, and the swing is large enough that the same site on the same network reports very different RPMs across a year without anything changing.
Most figures shared publicly come from good months, for the obvious reason that people report numbers they are pleased with. A December figure compared against your April figure is not a comparison at all.
The variables inside your control
Of everything above, the network is the smallest lever. These are larger, and all of them are yours:
- Pages per session. Revenue scales with impressions, not visitors. A site averaging four pages a visit has four times the inventory of one averaging one. This is the most common explanation for two sites with identical traffic and very different earnings.
- Ad density and placement. More units earn more per pageview and cost readers. Where that trade sits is a real decision, and it differs by category — a reference site can carry more than a site people read at leisure.
- Page speed. Ad code is heavy. Slower pages lose visitors before the ads render, which shows up as lower revenue that looks like a rate problem.
- Which pages get the traffic. A surge onto a thin page earns much less than steady traffic to deep content, so a viral month can lower your average RPM while raising your revenue.
How to compare properly
If you want a benchmark that means something, you need a site in your category, with a similar audience geography, a similar session depth, in the same month. That is a demanding set of conditions, which is precisely why most published comparisons do not meet them.
The practical alternative is to stop comparing outward and compare inward: your own RPM month over month, against your own history, with the seasonal pattern accounted for. It is the only comparison where you actually control the variables — and it is the one that will tell you whether a change you made worked.
What we can offer on the outward comparison is the part that is verifiable: which networks carry which kinds of publisher, and in which categories. That is in the approval-difficulty ranking and the per-category breakdowns. We will not publish RPM tables, because we cannot see anyone’s revenue and neither can anyone else who publishes them.
Common questions
What is a good RPM for a blog?
There is no answer that survives contact with the variables. The same site earns very different RPMs in December and April; two sites in different categories with different audience geographies differ by multiples at identical traffic. Any single figure quoted as "good" is describing one site, in one subject, in one month, with an audience you know nothing about.
Why is my RPM lower than what other bloggers report?
Most likely audience geography, category, session depth or the month — in roughly that order, and usually more than one at once. Reported figures also skew high because people publish numbers they are pleased with. Before concluding your network is underperforming, check your pages per session and where your readers actually are.
Do bigger ad networks pay higher RPMs?
Not reliably, and the comparison is usually confounded. Networks with a higher entry bar have publishers who were already stronger, in better categories, with more domestic audiences — so their reported figures are higher for reasons that precede the network. The publisher spread table above shows how differently composed these rosters are.
Why do Ezoic RPM reports vary more than Mediavine's?
Because the publisher bases differ enormously in consistency. The spread column above measures the gap between a network's weakest and strongest publishers, and the widest-funnel networks span several times the range of the selective ones. A network that accepts nearly everyone will produce scattered reports; a network with a high floor produces consistent ones, because its publishers resemble each other.
How can I increase my ad RPM?
The largest levers are yours rather than your network's: pages per session, ad density and placement, page speed, and which content the traffic lands on. Improving session depth in particular tends to move revenue more than switching network does, because it multiplies inventory rather than adjusting what each impression sells for.
We index advertising relationships, not revenue. Nothing here is a claim about what any network pays, because that is not publicly observable. Every figure on this page is read from our own index when the page is built, and refreshes on a weekly cycle — so the numbers move as the data does. Method and evidence rules: coverage & method.