Forum Moderators: martinibuster

Message Too Old, No Replies

My Experience with of AdSense CPM by Country

For those who are interested....

         

dataguy

8:00 pm on May 30, 2010 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Some AdSense publishers subscribe to the notion that by preventing lower-paying AdSense ads from being displayed, overall AdSense ad revenue increases. Despite Google's claim to the contrary, I am one of those who buy into this notion.

Besides removing AdSense from pages with a history of low AdSense CPM, I also prevent AdSense from being displayed to visitors from certain countries which have a history of producing low AdSense CPM. I try to adjust my filter at least once a year. This time I thought others might find my numbers useful.

The following list is a list of countries and the CPM that they have produced on my U.S. based website as a percentage of the CPM from U.S. based visitors. (CPM from U.S. based visitors = 100%)

These numbers are taken from a one month sampling using Google Analytics for an English language site based in the U.S. with about 100,000 pages covering a range of over 500 widely varied topics of user generated content. The site receives about 3 million unique visitors per month. Of course, this may or may not apply to your website depending on numerous factors, but I think it gives some pretty good insight into where the money is.


Country / % to U.S. CPM
Switzerland: 138%
United States: 100%
United Kingdom: 67%
Canada: 62%
Australia: 59%
Ireland: 48%
Sweden: 48%
Belgium: 46%
Lebanon: 42%
South Africa: 40%
Spain: 38%
Norway: 35%
Qatar: 34%
Denmark: 33%
Brazil: 29%
Russia: 28%
New Zealand: 28%
France: 27%
Saudi Arabia: 27%
Austria: 26%
Hong Kong: 26%
Bulgaria: 26%
Greece: 25%
United Arab Emirates: 23%
Singapore: 21%
Kenya: 21%
Netherlands: 20%
Colombia: 19%
Italy: 18%
Hungary: 18%
Puerto Rico: 18%
Iran: 17%
Bangladesh: 16%
Serbia: 16%
Turkey: 15%
China: 14%
Germany: 14%
Pakistan: 14%
Israel: 13%
Japan: 13%
Finland: 13%
Nigeria: 13%
India: 12%
Ghana: 12%
Jamaica: 12%
Vietnam: 11%
Malaysia: 11%
Mexico: 11%
Croatia: 11%
South Korea: 11%
Ukraine: 11%
Trinidad and Tobago: 10%
Indonesia: 10%
Poland: 9%
Argentina: 9%
Czech Republic: 8%
Romania: 7%
Kuwait: 7%
Philippines: 6%
Thailand: 5%
Portugal: 5%
Sri Lanka: 4%
Egypt: 4%

This list represents the top 63 referring countries to my website.

I hope you find this information useful.

IanCP

9:46 pm on May 30, 2010 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



Interesting and somewhat reflects my experience.
"by preventing lower-paying AdSense ads from being displayed, overall AdSense ad revenue increases"

I've yet to be convinced of that.

All too often with AdSense, because of the diversity of our sites, and the one million and one other factors involved, there is never a "one size fits all" solution or answer.

Everyone's individual experience with AdSense compared to others is like comparing apples with every other kind of fruit and vegetable known to mankind.

gouri

10:41 pm on May 30, 2010 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



The % to U.S. CPM that are below 50% - Is this because of smartpricing? Visitors to a website from these nations are not likely to buy products whose ads they click on so Google wants to reduce the Adsense payments?

dataguy

11:21 pm on May 30, 2010 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



I've yet to be convinced of that.

I've had what I considered incontrovertible proof of this at one time, though now I'm not as certain since interest-based ads have become available. It sure seems like removing ads from low paying pages increases revenue, but I can't say that I can prove it anymore.

The % to U.S. CPM that are below 50% - Is this because of smartpricing?

I don't think it's smart pricing, I think there is just less demand for ads being displayed in less affluent countries.

farmboy

1:42 pm on May 31, 2010 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



Belgium: 46%
Egypt: 4%


1. That gives the impression it would be 10 times better to work on your site to attract visitors from Belgium rather than Egypt. But suppose during the month in question only 2 people from Belgium visited and both clicked an ad. Wouldn't that distort what you might experience over an entire year?

If I get hired by a baseball team, get one at-bat and get a base hit, I'd have the best batting average in history. For that reason, you have to have a minimum number of at-bats before your record is considered.


2. This relies on Google being accurate in knowing where visitors come from. They sure seem to be confused about where I am located.


FarmBoy

netmeg

2:54 pm on May 31, 2010 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



So what's up with the Swiss?

farmboy

3:23 pm on May 31, 2010 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



So what's up with the Swiss?


Sven and his brother were the only two people who visited the site during the month and they both clicked lots of Russian bride ads?


FarmBoy

dataguy

3:28 pm on May 31, 2010 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



1. That gives the impression it would be 10 times better to work on your site to attract visitors from Belgium rather than Egypt. But suppose during the month in question only 2 people from Belgium visited and both clicked an ad. Wouldn't that distort what you might experience over an entire year?

Yes. For these numbers though, none of these countries had fewer than 5,000 visitors to my site. I wouldn't recommend finely tuning any web site for any demographic that could possibly only produce a handful of visitors in a month.

So what's up with the Swiss?

Good question. Maybe it's all the formerly U.S. based companies now located in Switzerland, with money to burn.

wanderingmind

10:40 pm on Jun 3, 2010 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



dataguy,

You were using Analytics data for this? My own findings match it pretty much.

Is it possible to track individual pages' revenues by country in Analytics though? Could not find a way...

JasonDX

9:39 pm on Jun 4, 2010 (gmt 0)

10+ Year Member



I also prevent AdSense from being displayed to visitors from certain countries which have a history of producing low AdSense CPM.


What script do you use for this? Can you recommend one? This something Google would probably never offer to publishers.

jetteroheller

12:15 pm on Jun 5, 2010 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



I just tried to do with Google Analytics the same statistic.

I did it some month ago.

But now I can not find out how to do it.

Please advice

dataguy

6:09 pm on Jun 5, 2010 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



What script do you use for this?

It's custom coded using IP data from IP2Country.

You were using Analytics data for this?

Yes, you have to have your AdSense account associated with your Analytics account. Then you have to do Content > Map Overlay and then do Top AdSense Content and for the secondary dimension (the drop-down box which says "none") select Country/Territory. Then you'll have to do an advanced filter for just the specific country that you want, and see the total in the top bar. It's not intuitive nor quick.

FWIW, my personal experience is that after I turned off all filtering for a month to get a baseline, Monday I turned this country filtering on, and Thursday and Friday of the same week revenue increased 27% over the Thursday and Friday of the previous week.

Chrispcritters

4:46 pm on Jun 8, 2010 (gmt 0)

10+ Year Member



I've often wondered about this same topic. Not so much that by not serving ads to traffic from specific countries might increase the eCPM rates of traffic from countries that pay well, but that there might be other somewhat more intangible benefits.

Let's assume the 80/20 rule. 80% of revenue comes from 20% of traffic. If I'm willing to take a 20% cut in revenue what happens to the experience of the 80% of traffic that no longer sees ads?

Would the visitor from the ad excluded country have a better user experience? Faster page loads, less distractions, more page views, more time on site, lower bounce rates? Would people be more willing to link to this site? More likely to otherwise mention the site to a friend?

Would these become positive signals to search engines and help site rankings thus increasing the traffic from countries where ads are served? Enough to overcome the 20% loss from before.

alsid

3:43 am on Jun 11, 2010 (gmt 0)

10+ Year Member



dataguy

and what about CPC, its relative (to US) or absolute mean and variance by country? Could you please show us some data if possible.

Of course it depends on content, pagerank, history for every page and website etc but still...

Surely your report would be really perfect if you show some such info.

dk82

4:39 pm on Jun 14, 2010 (gmt 0)

10+ Year Member



I also prevent AdSense from being displayed to visitors from certain countries which have a history of producing low AdSense CPM.


What script do you use for this? Can you recommend one? This something Google would probably never offer to publishers.


Geotargeting is also available in the free Doubleclick for Publishers (DFP, formerly Google Ad Manager) product. You can select cities, metro regions, states, countries, or continents.

For example, if you have several direct sale ads for particular regions, you can configure those accordingly and have AdSense automatically backfill for any "unsold" region. Conversly, if AdSense isn't performing well in certain regions, you can specify alternative ads (affiliate programs, etc.) in its place.

farmboy

6:58 pm on Jun 14, 2010 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



How much confidence do others have in Google's ability to know where a visitor is located? I have little confidence based on my own experience with my "fixed" computer.

With netbooks, wireless cards, phones that can access the web, etc., I really wonder how reliable the data?

Is Google really good at knowing whether a visitor is from northern Michigan vs. southern Canada, for example?

Or in Europe with a number of small countries close to each other.


FarmBoy