Forum Moderators: martinibuster
It was nice to get, but I wasn't experiencing any real challenges with the program. Is this to assure me all will be well, or preparing me for things to come?
The odds of me making nearly the same amount of money for four months straight, given all the variables involved, and my history with Adsense the past four years for each of those months has to be so rare if it were beef, I would have ecoli by now.
I think that the odds of any one publisher out of all their publishers to make the nearly the same amount of Adsense money for 4 months straight, without any kind of cap on their part, is probably very high. It would be like rolling a six 4 times in a dice row. Not likely if one person did it, but almost assured that some people would throw four 6s if you had a large enough people throwing a dice.
I think you'd have to have a lot more data, including historical earnings from a lot more publishers to really draw any conclusions that they had any kind of earnings cap.
With my own sites I personally don't see any kind of cap at all.
Excellent observations - thanks for sharing.
While I agree with JaneDoe that more data would make this excercise even more useful, I think you have really nailed it down.
Looking at my own statistics, I see the following:
- eCPM has dropped like a rock in September and again in October, now down to an all-time low
- Interestingly, EPC for October was the BEST in 2008. Huh?
- EPC has remained the same within fractions of a Cent for the four preceeding months
- Daily revenue down 1/3 from last years October
My take? Google determines the individual publishers' EPC not based on an auction but on some other method that is as opaque as it possibly can be. I have no doubt that they (Google) are adjusting publishers' earnings as they see fit.
I think each publisher has not a "cap" but a "target" earning. This target lies in the future, e.g. the algo determines: "This publisher shall reach $X per month by 1st March 2009, by when we will re-evaluate the value of this publisher". The reaching of the goal is ensured by adjusting EPC. In some rare cases, the target is reached prior to the target date, for example when the traffic changes very little, or when the publisher has been for a long time with Adsense.
Maybe they even just look at average eCPMs? When I look at my 200 day moving averages, I see a couple of "interesting" and abrupt changes. Something that is very very unusual when looking at 200 day averages:
a) Exactly 9 months into the program, the 200 day average for eCPM went from a steep growth (8% per month) to moderate growth (0-3% growth).
b) About 7 months later, the moderate growth went into decline mode (-3% to -5% per month). The plotted data almost looks like drawn with a ruler. Up to today I have no rational explanation for this steady decline.
c) Since February 2008 I seem to be in the "unchanged" mode, i.e. hovering around 0% growth rate for eCPM.
d) Since around mid-September I am again set up for a decline. Hence the mail?
Again, I do not think that the payout is purely based on the auction price/maximum bid. It seems to acknowledge a lot of other factors. The "target algo" ensures that changes happen just in so tiny fractions that they are invisible, even for the individual publisher. Rolled out across a multitude of publishers, everyone is seeing different data, so any discussion is futile.
Otherwise the massive volatility in daily data would not make sense: One day you see a eCPM of X, the next day three times as much? Com'on. This is the target algo fiddling around with the data to meet the (longterm) target.
(Now ducking for the fanboi wave.)
Click fraud is easy to avoid if you calculate a site's target earnings based on various characteristics.
By putting artificial limitations and delays on a sites earnings Google can guarantee advertisers are not being defrauded.
After a point, these types of limitations make attempting to optimize a site's Adsense performance a pure waste of time. (Being stupid I still keep trying!)
There may be certain sites that are far less prone to click fraud, perhaps a highly technical site versus a social site. Or a site targeted at a specific industrial vertical, billing perhaps? For these sites, perhaps the constraints are very loose so their earnings more closely track the bidding on Ad terms.
The other factor is Google's silly discounting, giving the highest bidder the second highest bidders price. This really suppresses earnings for publishers.
I wrote a long post bemoaning this approach as the actual reason Google needed smart pricing in the first place.
Smart Pricing caused the MFA craze and punished publishers.
Google has backed themselves into a bad system.
[webmasterworld.com...]
If you attempt to read this post I guarantee you will need a nap and have a headache too!
To be fair it could be part of this update, at the same time the email may have been fortelling this "update"
I've been seeing a very sluggish CTR this fall, even as average EPC has been climbing. That seems pretty reasonable, given the state of the economy: Readers are holding off on purchase decisions, and advertisers are competing harder than ever for the prospects who are looking to spend money right now. That's my hypothesis, anyway, and it makes more sense than the notion that Google is fiddling with the numbers (a claim that we hear very quarter, but which is never borne out by the available evidence).
I'm curious as to what the criteria was as to who got the letter and who didn't.
signor_john AND ALL
If you did not get this email please read this new Adsense thread.
Not receiving all your Adsene emails? [webmasterworld.com]
(Mind you, I suppose it's possible that the e-mail was sent to my nonexistent or inaccessible gmail address, even though other e-mails from AdSense have been sent to my default mysite.com address.)
As I review each Google service, I believe there are inconsistencies in just about all of them. Adwords seems to have just thrown away my old email address altogether. I can find no sign of it.
I guess this really only affects "old timers". Perhaps 2006 and before?
Oh well.
I guess this really only affects "old timers". Perhaps 2006 and before?
Dunno, but I've been using AdSense since the program launched back in 2003.
On the other hand, I haven't used gmail for anything but a quick test or two in its beta days.
dunno, borne, frontpage, 2003, travel, europe
signor_john = efv ?
...in its beta days.
No mail for me from AdSense.
