Forum Moderators: martinibuster
I came up with a theory that smart pricing was not really all that smart and tried to figure out what the bot does to determine your "smart pricing"
I think that there is an algo that ties in your poor performing pages and uses them to weight your site for worthiness.
The more pages you track the lower your quality (to the bot).
On the 2nd I deleted all custom channels and saw a 400% jump in income. Went through the 4th of July weekend with better than average earnings. Daily income going up by 2 to 8 dollars a day.
Ctr and ecpm are doing better but slowly, with ecpm rocking down and up.
Payout per click went up from 3 to 9 cents and had 5 .19 cent clicks on search...unheard of in the past. (my average payout per click was 8 cents for my niche)
Noticed better ads began showing up that had not been there before and no crap ads that I could find.
Things slowed down and started rocking so I thought, "hummm, mr bot has found another way". So I started going though my site page by page and changing out the channel ads for regular ones and adding them to pages that did not have them...more page views...:), and have seen the ctr and epcm renew their upward creep.
This weekend was higher than the weekends I've been having for several months now.
For me it is working, I still have over 800 pages left to examine and check for adsense channel ads, (doing this in alphabetical order.)
I am not saying anyone else should do this, that is your choice. All I am saying is it is working for me and judging by the payout and the better ads I can tell the worth of my sites have gone up.
Doing the volume we do we would never even think about implementing this unless we heard of 90-180 days of solid impressive growth.
We have been with Adsense from the start and have 8 months of non channel data and I cannot see any decrease or correlation with or without channels.
I am not knocking the idea at all and most likely Ann is on to something but remember, this could be a double edged sword with smart pricing. Who knows what long term ramifications this could have on your account since we do not know what smart pricing takes into consideration.
What if smart pricing determines in 90 or 180 days that your account is not converting as well due to the loss of channel data and you drop 90% in epc?
Now if you have nothing to risk and are producing a small amount of income the risk is probably worth it but if you are pushing 6-7 figures per year I would strongly think twice about this.
The one thing that I found out is Adsense likes slow moves.
So don't put all your encased potential chicken embryos in one circular woven transport receptacle, especially if your a big bird.
As you may recall Ann, I got this confused awhile back too: it's not Smart Pricing that hasn't been confirmed/denied by Google - it's the Sandbox (having to do with indexing of new pages in their engine) that they haven't copped to one way or the other.
Slowing down for the weekend. I sincerely hope this lasts but it may not. Only time will tell as said in the above post.
REMEMBER--I didn't tell any of you what to do, ONLY what I was doing.
There could be many unknown factors at work here so what works for me may not work for others....I believe I said that before :)
Good luck.
Ann
That simply wasn't the procedure Ann laid out!
BTW, Ann... I'm in as well. "Now vee vait"!
Chapman
I hope the results are positive for the testers. If this pans out, Google just may rework things soon...or removing just the "channel" part of the equation from the metrics. I would think they go by URL and other such metrics too but you never know what they are up to for sure.
Too bad there aren't API calls for channel creation/deletion (or is there, cannot recall, nah, just one for AW), then you could create fresh channels each morning and purge them each night (after analyzing your channel stats for the day). Oops...they won't implement that now...
Will be ever know for sure, though? Could be coincidence, fluke, etc.
SB
[edited by: StephenBauer at 7:41 pm (utc) on July 21, 2006]
Frankly I can't blame mt1955. It seems way too early for anything to have affected smart pricing, but if my earnings had dropped 95% I'd probably bail on the experiment too.
The trouble with the assertion that removing channels will lead to this is that maybe we are mistaking correlation with causation. Without knowing more of the way the SP algo works, the knowledge we have to these sorts of changes is very speculative.
I found something that worked for me and shared it with the rest...In the past I have experimented with every way possible to try to improve my income and reverse the downward trend. And YES, I tried different ways with the channels!
This may be just a fluke, etc. I am not trying to start a "cult", far from it, it is simply a formula I worked out...take or leave it You are a singular individual with ideas of your own and you, of couse, must follow your own way.
But, substituting sugar for flour will not make a good gravy...sweet though:)
Ann
The last poster reminds me of the people who beg me for my barbeque sauce recipe. They ultimately substitute all or most of the specific ingredients and inform me at our next meeting that it wasn't very good! ;-)That simply wasn't the procedure Ann laid out!
Ok, :-) you got me fair and square... but I was also too chicken to delete them so I tried just disabling them... maybe my post would have been more useful to the thread had I written instead something like:
I meant no disrespect... just got me think'n about BBQ sauce! ;-)
I certainly can understand your hesitation to delete channels... I voiced some concern very early in this discussion as well. I do believe, however, that even if you DO delete them, you can recreate them (with the exact name) and pick back up where you left off! You would lose that recording period while they were "turned off" though.
I tested it this morning with channels I had deleted months ago and it appears to work.
Chapman
Things seemed to have returned to normal. But actually it's not like me to do anything that drastic. I usually just make small incremental changes, mostly page additions.
I wonder though, how much of what we do really has an effect and how much is just the normal variation in the business. The last 24 hours could have tanked for my sites regardless of the channels being disabled/enabled. Why not? It is the middle of the summer on a Friday when thunderstorms knocked out power on the plains, in the midwest and on the east coast. Today probably should have tanked.
Another example - early on I spent a lot of effort tweaking my pages and playing with the google_ad_section tags trying to get the ads to be more on topic. Then I gave up. Then I noticed sometimes the ads seemed to fit in with the content better, other times they didn't. It occurred to me that the G-machine was probably just serving up the best matched ads available at the moment and sometimes it could match them to the content better than other times.
So now I believe that my results are simply better when the content and ads match up pretty well and the bids for those ads are good. My results are not so good when there are not a lot of high paying ads lined up for my kind of content... and that I can probably make things worse but, other than adding content and getting more traffic, can't really do much to make the results a whole lot better.
Having said all that - I am still going to watch this thread with interest and thanks, ann for sharing the results of your experiment.
Can I have the recipe for your super duper world famous amazingly incredible etc Barbecue Sauce?
Pretty please? :)
Crystal Pegasus.
I have three sites, the flagship is over a thousand hand made pages that are BG (before Google), and one that originated about the time Google came into being, maybe before? and one completed a few months back.
this is just for your information since you felt you had to shout it out....just thought I would give you the low down on my sites since I LOVE to hear the truth shouted out, don't you?
I would really love it too if all those who did the exact same test I did would post in this thread about your results in about a week--or week and a half just to get some feedback on whether or not it is a viable option.
Another good day and fridays had really gotten bad lately. :)
Ann
Anyway I'm sure there will be no shortage of anecdotal evidence coming in over the next few weeks.
Chapman wrote:I certainly can understand your hesitation to delete channels... I voiced some concern very early in this discussion as well. I do believe, however, that even if you DO delete them, you can recreate them (with the exact name) and pick back up where you left off! You would lose that recording period while they were "turned off" though.
I tested it this morning with channels I had deleted months ago and it appears to work.
I've also tested the recovering of removed channels with their old data, and in my testing it only works for URL channels (as Google says), and not for custom channels.
In the custom case, a new channel is created with the same name and a different number. The data of the removed channel (removed from the user interface) is in fact not deleted but inaccessible, excepting if you keep an old form on your browser containing the removed channel, as I mentioned.
With that old form, for example some days later, we can make a variety of different queries for the removed channel, to verify that it is not any cache but real data that is in fact not removed. And this is indeed the case.
This is why I think that it's probably unnecessary to remove channels, and that deactivating to stop collecting new data would be enough for this testing, unless we really don't need the old custom channel data. In this case, of course we can go ahead removing. But this would be hard for many publishers; and if they just deactivate probably there is not difference with removing for this test.
I think this experiment is interesting even if we later find that there is not a clear correlation between using custom channels and the profit increase Ann is experiencing. In my opinion there is a higher probability that this improvement is being caused by another unknown reason, but I can be wrong, naturally.
In any case, the result will be useful knowledge, and as the classic quote on testing says:
"Why, I have not failed. I have just found 10,000 ways that won’t work."
(Edison, before discovering the nickel-iron battery)
Thanks... I stand corrected!
The channels I reconnected to that had been deleted months earlier were indeed URL channels (they were the only ones I had previously deleted).
The appearance that I could delete a custom channel and still access the data was, as you've illustrated, simply because I was still was maintaining a browser session that had the neccessary name/number association.
Sorry to be unintentionally misleading.
Chapman
I haven't tried this but it a possibility.
Ann
Backing up a copy of each page with the ad on it that you were tracking when you deleted the custom channel then if you want it back create another custom channel with the exact name and re-upload the page. might do the trick.
Couldn't you just copy and paste the html code for the channeled ads you want to delete and save it on your computer, then if you decide to create the exact same channel again, you'll have the exact same channel number and use the same name. Seems to me that would be the exact same channel and you can recreate it that way.
In fact, I wouldn't recommend the following for many publishers who need their custom channel data, but I'm doing this test with Ann's complete version, that is all active and inactive custom channels -and some old URL channels- removed from the user interface and the site pages, because anyway I'm using currently almost 200 URL channels, and -until two days ago- only a few unnecessary custom channels, just removed. My channel lists were cluttered with too many tests from the past two years, that were extremely useful but I don't need anymore.
Only URL channels are very useful for me currently; in fact 200 URL channels is a very low number.
So, just in case, I saved many queries on CSV for those old channels, and removed them to clean the lists, and at the same time to do this test, even if I would be surprised if it gave positive results for this thread's theory.
These two days I'm seeing the typical ups and downs, nothing conclusive. And two days and one or two people's data aren't enough, naturally.
So, probably it's unnecessary to say but, before removing your custom channels, please be sure that you don't need your old data. If this is the case, it's easy for you to help in this testing.
In other case, if you need your data but wish to do some testing, as also said, I think you may try just deactivating your custom channels for a week (or more if your traffic is low) to see the effect -or no effect- of stopping to collect new custom channel data.
21st showed higher Earnings per click, and today (22nd) as well. But CTR has been below average. Ergo: overall earnings the same as two days before with the difference that I have no channels any more :(
So I can't recommend this method. But still thanks to Ann for sharing your experience.
I believe that it is simply a big coincidence that your earnings have improved. Could this be?
Good luck to everyone!
It is much too soon for your test to have shown much. Higher paying clicks is only the beginning, it does show you are starting to get higher paying ads. Give it a chance since you have gone this far. :)
My ctr and ecpm is still on the low side but it is rocking up one day down the next, higher than the up day, etc, and gradually trending upward, hence it being too soon to judge.
I was down to less than a dollar eCPM when I started this test and now it is over 2 and still rocking. (CTR is doing the same but it was "only" down to 2.00 now hitting 3 and rocking)
That is why I was so upset! Previously high earner with no known action for pulling a quality site down. My daily ave income has jumped up significately.
I would rather have a slow steady growth than a fast track that runs out too soon like the other things I have tried.
Funny thing is: very few crap ads, and about the same amount of clicks I have been getting. Shows that better ads='mo money :)
Good luck,
Ann
Without making any changes to my site, over the last couple of weeks since QS I've seen CTR average rise by a couple of percent, I've seen the daily average epc go up by as much as 20c per click over the month's average before QS. Mind you, the first couple of weeks of this month were pretty ropey compared to the month before, so even those changes may not be due to QS, but simply a ropey couple of weeks. I've seen fundamental changes in the ads that show on my site, and Google's search pages for my search terms.
So whilst you are all busy trying to assess the effects of this idea, please bear in mind that some of what you are seeing might be due to general shifting of tectonic plates since QS was introduced.
But I still don't understand why Google would have any interest in showing you intentionally lower paying ads instead of higher paying ads - can you elaborate on that a little bit more?
[edited by: HuhuFruFru at 7:56 am (utc) on July 23, 2006]
The reason for the lower paying ads, in my theory, is the flawed algo may have caused the system to view your site (and Mine) as low quality because of the way the algo seems to be flawed, using the custom channels info to help weight or rate the site. Therefore, they would reserve the better paying ads for what the Algo deemed to be a higher level property and would send the higher paying ads to others they deemed good quality so the advertiser, the one who paid more for the ads, gets more bang for their bucks.
But since the whole thing seems to have gotten out of whack the good sites were given the lower paid ads just because they were doing due diligence in testing out new pages or lower performing pages to try to make their site more profitable for both ourselves and Google.
Just my opinion.
Ann
Ann wrote:(...) because of the way the algo seems to be flawed, using the custom channels info to help weight or rate the site. (...) testing out new pages or lower performing pages (...)
And all this without taking into account other factors like the one mentioned by David_uk.
So, even if this theory is right, the test results might be inconclusive.
