Forum Moderators: martinibuster
In any case, October glitch or not, have you ever been smartpriced (i.e. had drastically decreased revenues for no apparent reason) and if so, how long did it take for your earnings to start improving (if at all)?
- I don't think there's any way to know if you've been "smart priced," since earnings per click can go down for any number of reasons.
- Not all earnings drops are caused by smart pricing, or even by a drop in average earnings per click. (Some people here have complained of drops in clickthrough rates, which are a whole separate issue from smart pricing.)
EFV, in our case CTR has remained untouched since our earnings dropped but what we find peculiar is, irrespective of CTR, our eCPM falls and rises in order to meet an almost static daily earnings amount. We could have a record-breaking day of traffic and record breaking CTR (which we just did, coincidentally) and our eCPM would dwindle extremely low to keep us pegged at the same income. If traffic decreases the following day with a decrease in CTR to average levels, eCPM will increase to -- you guessed it -- end the day at the exact same income.
It's disheartening to watch your CTR rise through increased traffic flow while watching eCPM rocket south to keep you pegged at static daily income.
[edited by: CentennialEmpire at 1:04 am (utc) on Jan. 16, 2008]
It's like Google is playing games.
I keep thinking that, but why would Google want to rip us off?
Just an opinion...
Perhaps Google's tactic is to artificially decrease a publishers earnings in the hopes that a publisher will advertise with Adwords?
1) But why that publisher, and not another publisher?
2) What evidence do you have that AdWords/AdSense sales are declining? The 4Q 2007 earnings report isn't out yet, but in previous quarters, content-network revenues and publisher payments have increased.
3) Since the topic of this thread is (or was) smart pricing, it's worth noting that Google earns less from "smartpriced" (a.k.a. discounted) clicks than from clicks that are sold at full retail. If Google wanted to pad its profits at the expense of publishers, smart pricing wouldn't be the way to accomplish that.
- I didn't say one publisher was affected. In the first post I specifically said I'm basing my assumption on multiple accounts of publishers seeing substantially decreased revenues in what appears to be concerted roll-outs of "smartpricing" or "something else affecting earnings." One group experienced a decrease in October, aka the "glitch," and another in December, with the first group still hurting. Who's next? I can't say, but there could be another wave.
- When it comes to "evidence," unless you work for Google and are exposed to the operations of the company, I can say with certainty that I have as much evidence to support my opinion as you have to counteract it. Adsense and "evidence" are like water and oil.
My desire here is to learn how and if publishers have recovered from substantial earnings slumps, aka "smartpricing," aka "whatever Google's up to." I'm not impressed that this thread is already deviating into a "what evidence do you have" fistacuff.
My desire here is to learn how and if publishers have recovered from substantial earnings slumps, aka "smartpricing," aka "whatever Google's up to." I'm not impressed that this thread is already deviating into a "what evidence do you have" fistacuff.
You insist on using "smart pricing" as a synonym for "lower earnings," when it actually refers to discounts for advertisers:
[adwords.google.com...]
Smart pricing could be one of the reasons why some publishers have experienced substantial earnings declines, but only if their average EPC has dropped sharply.
Trade that traffic for high quality targeted traffic that clicks and you might well see the upside of smart pricing after a while.
How often does it adjust? Speculation here in the past has been that it may adjust once a week. But that's just speculation.
But not every fluctuation in epc is smart pricing, sometimes it's just lower bids by advertisers, or maybe a bigger spender pulls out and overall epc crashes. And telling the causes apart is almost, if not, impossible.
What do you think is causing the "cap" we're experiencing, where no matter what the traffic and CTR is, eCPM decreases to end the day at the same value as the day before and the day before that?
[edited by: CentennialEmpire at 2:46 am (utc) on Jan. 16, 2008]
... in our case CTR has remained untouched since our earnings dropped but what we find peculiar is, irrespective of CTR, our eCPM falls and rises in order to meet an almost static daily earnings amount. We could have a record-breaking day of traffic and record breaking CTR (which we just did, coincidentally) and our eCPM would dwindle extremely low to keep us pegged at the same income. If traffic decreases the following day with a decrease in CTR to average levels, eCPM will increase to -- you guessed it -- end the day at the exact same income.
Which is why after almost 3 years, despite everyone telling me there isn't, I still believe that Adsense imposes an earnings cap and this cap (ceiling) moves up periodically :
[webmasterworld.com...]
If it's not a purposeful cap it's certainly raising eyebrows around here as we finish each day with virtually the same earnings, irrespective of CTR or traffic. If there's any deviation then earnings will dip, but they certainly won't rise.
In essence we're experiencing significantly decreased earnings and no matter what happens (spurt in traffic, spurt in CTR, etc), we can't earn any more than the "cap."
Since the topic of this thread is (or was) smart pricing, it's worth noting that Google earns less from "smartpriced" (a.k.a. discounted) clicks than from clicks that are sold at full retail.
How do you know? Are you an accountant at Google?
Has anybody read their last 10-Q report? Dated November 7. Read the paragraph "Trends in Our Business ". Read between the lines in the third paragraph. Ticker GOOG BTW.
If Google is charging the same but keeping a larger portion of the revenue for itself, that may be happening but by definition it is not Smart Pricing.
An interesting thread, and probably the right place to question some of the myths that are being published around here.
earnings per click can go down for any number of reasons
Yes, in an auction based system, prices may vary to a certain extent. BUT several members of this forum have experienced that Adsense EPC on any given day may suddenly drop to, like, 50% of the average values of the surrounding days. It's really like 100-103-97-114-60-105-123 when the first day is set to '100'. In a big sector, with advertisers being the same, with traffic and clicks also being in the '100' range, there is NO rational explanation for this.
To me (to me!) this looks like a money grab.
in our case CTR has remained untouched since our earnings dropped but what we find peculiar is, irrespective of CTR, our eCPM falls and rises in order to meet an almost static daily earnings amount
Indeed, the variance on eCPM is unnatural, almost random. In my view, this can not be explained by rational reasons.
But why that publisher, and not another publisher?
That's the question. But then again, why does all problems need to apply to all publishers. Wouldn't it be smart if you exposed only a fraction of publishers to any problem? This would prevent any insights into Google's Adsense blackbox. Which is exactly what they want. They DO NOT WANT TO LET ANYONE UNDERSTAND THEIR ALGOS. You can only reach this by making any attempts of analysis as useless as possible. So, "Group A" gets a new Smartpricing, "Group B" does not get it. Easy. How do I get to "Group B"? I don't know. Maybe write Google-friendly messages in certain message boards? ;-)
Google earns less from "smartpriced" (a.k.a. discounted) clicks than from clicks that are sold at full retail. If Google wanted to pad its profits at the expense of publishers, smart pricing wouldn't be the way to accomplish that.
SmartPricing would be the perfect way to do it. So, please, show us the evidence that Google is indeed giving mony back to advertisers. Without this evidence, all we have is Google's word for it. Which I do not trust at all.
So, "Group A" gets a new Smartpricing, "Group B" does not get it.
I don't think there's any way to know if you've been "smart priced," since earnings per click can go down for any number of reasons.
+1
Okay, for everyone complaining about smartpricing, let's hear how you'd code the algo if you were at the 'plex...
To the T/S: probably and I'm not sure.
Google has revealed about as much of its s/p algo as it has about its SERP algo.
Join the fog!
p/g
That's the question. But then again, why does all problems need to apply to all publishers. Wouldn't it be smart if you exposed only a fraction of publishers to any problem?
Why do you assume that it's a problem? In another thread, someone complained that AdSense was "broken" because his earnings had dropped massively. That's a typical example of "It's all about me" thinking. In reality, the system may be performing exactly as it should. Maybe Google wants certain type of content, certain topics, certain ad-usage patterns, etc. to pay less than others: for the good of advertisers, for the good of Google, for the good of the Web, or (just as likely) all three. Combine that with the an auction-driven, real-time marketplace, and you have a recipe for volatility in general plus occasional dramatic changes for publishers who meet certain Google-defined criteria.
When it comes to reduced earnings they appear to affect chunks of publishers within the same timeframe, also a group-a or group-b scenario but we have a much harder time understanding the discrepancies because of Google's secretive operation.
why would it be a conspiracy?
I think the term conspiracy refers to the fact that Google officially claims to "do no evil". According to this claim, some believe that whatever Google says or does must be honest and in the sole interest of the advertisers, users, publishers, the web community, and the world in general. And to those, even the thought that Google is a business that uses its unique position to squeeze every penny out of their customers and partners is already a conspiracy - in their eyes, Google can't do wrong.
Which is rubbish. I believe that Google leverages their market power (i.e. the weakness of their competitors!) and the fact that they can not be audited to increase their profit. And yes, there is nothing wrong with that. It would be nice, though, if they told the world about it instead of keep brainwashing everyone about "do no evil".
That is effectively a group-a and group-b scenario, no?
The "little arrows" ads are part of a test, and there isn't anything secretive about it: Google has discussed the test openly (here [groups.google.com], for example).
FWIW, threads about big drops in earnings have been a staple of this forum for the last several years: at least since smart pricing was introduced in 2004, and possibly longer. Sometimes the publishers who are affected blame smart pricing; sometimes they claim that Google has cut the overall payout percentage; sometimes they insist that advertisers are leaving AdWords/AdSense in droves. Very few have been willing to concede that Google might be targeting certain publisher or site profiles--something that would be in keeping with Google's use of profiling in AdWords and Google Search.
Perhaps the recent string of decreasing revenues reported by a number of individuals on this forum, including myself, have something to do with new smartpricing algorithms being launched across the network in chunks (which may explain why some sites were hit earlier [i.e. October glitch], some only recently, and others not yet).
I'm wondering what you think about the principles of parsimony and occam's razor?
There are an infinite amount of ways to explain anything, and you and others are certainly not choosing the simplest and most parsimonious.
..not to mention you really have no reliable data, and will not get sufficient reliable data this way.
1) NONE of us have any way of knowing if we've been smartpriced.
2) Better to talk about the actually data symptoms if you want to have a useful discussion (ie. sudden changes in earning, whatever) as in "How many of you experienced drops in income on such and such dates".
3) Sorry to say this but as useful as it "seems" to be asking your questions, the end result is conclusions that are no more reliable than flipping a coin to choose from a number of explanations. The problem here, and its a huge one, is not that we don't know, but that some people REFUSE to acknowledge that they don't know, and can't know.
These threads, (in fact most adsense threads) should be labelled "for entertainment only", because to act on ANY "data" posted here (it's not real data), and to make decisions based on bad data, and bad reasoning is destructive.
Some people know that. Newbies should be aware of that.
Very few have been willing to concede that Google might be targeting certain publisher or site profiles
I don't think it is 'the' issue as often as it is referenced here.
Has anyone ever tried to test if smartpricing is an issue by taking off one account's ads and replacing it with another to see if earnings are impacted. My understanding is that smartpricing is associated with the account and not site, so I suspect that changing to someone else's account would at least have a short term impact on improving the earnings if smartpricing was the issue.
Has anyone ever tried to test if smartpricing is an issue by taking off one account's ads and replacing it with another to see if earnings are impacted.
I have implicitly tested smart pricing with my business model. I currently have several hundred direct advertisers. Most of my directs (95%) say they love the performance of their ads and are renewing. My adsense earning as with many folks has decreased. If I have been smart priced, then the advertisers on Adwords are not seeing the same performance my direct advertisers are or the performance is not to Goggles par.
So, with all of this said something doesn’t fit with respect to my site and smart pricing. Who knows? Maybe nobody has been smart priced (by definition).
[edited by: Edge at 12:12 am (utc) on Jan. 17, 2008]
Very few have been willing to concede that Google might be targeting certain publisher or site profiles
I suspect that may be true. My best guess is that they may also look at some other quality factors above and beyond conversion rates in calculating smart pricing.
Back when adlogger could actually show you the target that people clicked on, I experienced what most people would call smart pricing or "the cap" or any of a number of other theories of the day.
What adlogger actually showed me was three new ads that had shown up, that were getting lots of clicks, but weren't paying much per click. One was a diet tea, another was a "become a secret shopper" and the third was "free laptop". They got their high position in the ads because of their high CTR, even though they were low paying. I stopped getting my higher priced clicks because they weren't being shown very often.
So there is one way that a higher CTR can mean just about the same eCPM.
As for getting more impressions, many traffic spikes produce lower quality clicks (digg, slashdot, etc) so it would not surprise me if they were discounted at first to revalue the traffic. I guess that would be a form of smartpricing, but if your traffic pans out, you should get your boost in income after a while.
Increases in traffic should also change your mix of ads. Some of the most on-target ads on my sites are small players with limited budgets. If my traffic increased by an order of magnitude in a month, I doubt I would get a lot more of their impressions, I'd get more of the large sites that advertise for every keyword under the sun.
It would be wonderful if earnings had a perfect correlation with impressions, but there are too many ways that it can get messed up, and most of those ways have nothing to do with smart pricing.
