Forum Moderators: martinibuster
I guess most of you would say my account has been smart-priced. Now what would be a good strategy to fight smart-pricing? Should I remove Adsense code from pages with low ctr? What else can I do?
My earnings are dropping again today...
I know there are skeptics that Smart Pricing takes CTR into account. However, Google has never explicitly denied it, and I'm inclined to believe that at least some of the time, something related to CTR DOES have an impact on Smart Pricing.
If you have channels that get no clicks, or only a few low-value clicks, you've got nothing to lose anyway. Replace those ads with something that might do better, and see what happens to your EPC.
Last night I removed 5 low converting sites from Adsense - and plan on removing more poor converting channels over the next few days. I'm thinking, my earnings are so low right now that removing these cant possible do any more harm (hopefully).
good luck
I know there are skeptics that Smart Pricing takes CTR into account. However, Google has never explicitly denied it
There is some confusing information in Google's booklet about this. It says
"don't restrict your revenue potential by limiting your ad placement to your site's most popular pages. With Adsense, it's the click through rate that counts, not number of page views" (p15)One can read this either way: the first part of the sentence suggesting CTR doesn't matter, the second that it does.
If ASA is reading this, it would be helpful to have some clarification of the statement on page 15. Is the official recommendation to publishers to keep ads on pages that have low CTR?
Yesterday was my 9th best day ever and that was on Superbowl Sunday and today it isn't even noon and my income is skyrocketing to what I can predict will be my best day ever - it's gonna be a killer month.
Anyway, my point is I never thrash at earnings changes like yanking AdSense from pages, so the low converting pages are still low converting, and income is spiralling upwards since Jan 1. Compare this to those that yank adsense from sites/pages when they claim smart pricing is beating them up, may have some impact but I'm not convinced as I let it ride and it went back up better than ever.
If anything both parts of that sentence are saying the same thing, that CTR is important so don't just put adsense on pages that have a lot of pageviews.
I guess quoting something out of context can give a misleading impression.
Page 15 is about "ad coverage", and the whole of the page basically says 'put your ads on as many pages as you can'. For example, the summary at the bottom has the heading:
The equation: More ads = More revenue
In context, therefore, saying "it's the clickthrough rate that counts" seems contrary to the rest of the page which encourages prolific use of ads.
It is an important point to get clarification for, though. We, like many publishers I guess, have a small number of pages that distort our CTR rates. We could double CTR by removing ads from those pages, and only lose a small percentage of our revenue directly. But when we have done this in the past, although CTR and eCPM go up significantly, earnings go down.
21, so you have experienced that removing low performing pages has hurt your account?
Most forum members believe the exact opposite is true.
Oh dear, how now brown cow?
(Thought it would be nice to show you a poem for a change)
They move money I control inventory and amount of impressions given to them, here's what I do:
a) Keep eye sharply on eCPM and not earnings for long term gains
b) Switched some of the low eCPM inventory to other networks.
c) Removed a high impressions 728 x 15 low ctr low eCPM adlinks channel.In effect I am restricting Google to the pages that perform best with Google, they are now getting 50% less inventory and earnings are down by only 20%, when earnings rise again with half inventory to full level, I do nothing, until the decline in earnings cycle start again, then I start raising the inventory to compensate, but with higher eCPM this time, once in full inventory and full earnings I let it sail, if eCPM drops I repeat by lowering inventory ..
It is very similar to sailing against the wind, controlling the size of your sail, and going in a zigzag pattern.
21_blue:
Is the official recommendation to publishers to keep ads on pages that have low CTR
I hear what you're selling but I'm not buying yet as AdSense is definitely a cyclic thing.
Are you BTW implying there is a smart-pricing cycle, meaning things will or might recover even if I do nothing? And if so, how long does the smart-pricing cycle take? Weeks or months?
H.
how do you know any of your changes make a difference and it's just cycles of ad budgets making all the difference instead?
Good question incrediBILL,
I only know it works, it took time to reach this pattern, and I also did isolation tests where I let it sail without intervention, and it only sinks more. (My period for each test was one week)
How does it work? Why does it work? I could care less, I am just happy it does and for now I am able to salvage my eCPM from the drain when things get bad. Will the same way always work on the long run? I don't think so.
There is definitely a cycle, many like me observed a Thursday update, my way does not define a cycle, it's more like breaking it when it gets vicious.
Anyway I am the last person on earth to claim understanding how SmartPricing works!
As I said earlier, it works SOME of the time to take ads off the lowest-performing pages. You can find out if it works for you only by doing it.
Smart Pricing, we have been told, does take conversions into account. However, not all advertisers do conversion tracking. That could explain the different results. On my site, maybe very few advertisers do conversion tracking, and the smart pricing algorithm falls back on something related to CTR. On another site, most of the advertisers DO use conversion tracking, meaning that smart pricing ignores CTR.
Start only with the very lowest-performing channels. And see what happens to your overall earnings after a week or two.
To look at things in the long term, from google's perspective:
If they believe a site is generating clicks but that the those clicks are not converting, yes, it would be in their interest to cut the payment to that site and refund money to the advertisers so that get better value and keep placing ads.
BUT, if they are going to penalise a whole account because some parts of it are not performing well, they risk losing the publisher to another ad provider.
HOWEVER, this is only the case is they keep the reasons for penalising sites a secret. If they believe that low CTR is indicative of poor conversion rates, surely it would be in their long term interests to ensure that their publishers know this and take appropriate action so that they are not tempted to leave google because of poor earnings.
I just cannot see any long term gain on google's part from being secretive about it.
This does not imply that there are not some other smart pricing algorithms that they would wish to keep secret - I just can't see a reason why this one should be, because making it well known would simply cause publishers to omit the adverts that google (allegedy) don't like.
Having read this thread, and some other similar material, I would like to ask those who think that CTR is a factor in Adsense smart pricing, just what they believe google's motives would be in penalising people for some activity, but not telling them that they would prefer that they refrain from said activity.
Moglex, your analysis makes a lot of sense. But it could be applied to other aspects of the AdSense system too, such as how the algorithm decides to serve up specific ads. Why, then, doesn't Google tell us?
You would think that it would be in Google's interest to tell us a number of things that they don't. But the reason they don't is that they want AdSense's inner workings to be as much a black box as possible to us, to keep people from gaming the system. That's my explanation, anyway.
In effect I am restricting Google to the pages that perform best with Google
Another valid strategy might to restrict AdSense to the pages that perform worst with AdSense, on the theory that those pages aren't likely to perform well with other revenue sources (such as affiliate sales) and you might as well earn whatever you can get from them with AdSense.
In any case, I agree with comments made elsewhere by 21_blue, who has pointed out that attempts to make eCPM look better may result in lower total earnings. (And earnings, not eCPM or CTR or EPC, are what you take to the bank.)
In any case, I agree with comments made elsewhere by 21_blue, who has pointed out that attempts to make eCPM look better may result in lower total earnings. (And earnings, not eCPM or CTR or EPC, are what you take to the bank.)
Which is why I said to look at earnings, not at eCPM, in my earlier posts. If you experiment with removing low-performing pages or channels from AdSense, it's a success only if your actual earnings go up. They have for me, quite markedly in one case, less so in two other cases.
Another valid strategy might to restrict AdSense to the pages that perform worst with AdSense
those pages aren't likely to perform well with other revenue sources (such as affiliate sales)
Another proof that sometimes less is more:
Yesterday I switched a 6 month old Leaderboard channel to a 468 banner size, and for 2 days this channel is producing Exellent EPC (50% better) and higher contribution to the site earnings, too early to make conclusions anyway but it does work, right now I am giving Google only 50% of my inventory and I am earning last month's full inventory daily average with a 30% boost in my eCPM.
And lower your eCPM and earnings.
eCPM and earnings = apples and oranges
Any earnings those pages make from other networks is good while you sanbox them, the $ diffrence is more than made up by the rise in AdSense earnings.
We've heard conflicting reports on that hypothesis. In my own experience, taking AdSense ads off lower-performing pages does not lead to higher eCPM and/or earnings.
Another proof that sometimes less is more:
Yesterday I switched a 6 month old Leaderboard channel to a 468 banner size...
Now you're talking about apples and oranges and pears. :-)
I would like to ask those who think that CTR is a factor in Adsense smart pricing, just what they believe google's motives would be in penalising people for some activity
CTR is often correlated with poorly-qualified traffic. It is not causal, however, so there are many counter-examples. Thus, Google will correctly tell you that you should not remove AdSense from low-CTR pages, and individual webmasters will correctly tell you that they improved their SmartPricing performance by doing just that.
Use low CTR to identify the low hanging fruit of poorly qualified traffic. But use your brain to avoid removing AdSense from low-CTR pages that are providing reasonably qualified traffic to AdWords advertisers.
I started removing ads from underperforming pages, reducing the number of ads per page on the remaining pages (along with weeding MFAs) some months ago and I havent looked back since.
Which is why I said to look at earnings, not at eCPM, in my earlier posts. If you experiment with removing low-performing pages or channels from AdSense, it's a success only if your actual earnings go up.
It’s not that simple. Earnings can go up for other reasons. I don't use raw 'earnings' as the sole benchmark for assessing changes made to a site.
Scurramunga wrote:
I started removing ads from underperforming pages, reducing the number of ads per page on the remaining pages (along with weeding MFAs) some months ago and I havent looked back since.
Looking only at eCPM can be misleading. But looking only at earnings can also be misleading.
Ah yes I hear what it is that you are saying.
However, although I say I havent looked back I did try restoring some of the underperforming pages at times whilst trying to focus the content more. On and off I put them back on and take them off as soon as things seem go downhill. Same with a couple of the borderline MFA's.
After all this, I have still come to the conclusion that only a core number of pages are doing exceptionally well, my ctr and weekly revenue has been rising as I have been focusing my content and yes, the increase did start before I embarked on content optimisation.
There are just those pages that don't work well no matter how much optimisation I try.
[edited by: Scurramunga at 1:41 am (utc) on Feb. 7, 2006]
--more ads increase ad server load
--more ads eat more bandwidth
--more ads mean more pages that the ad bots need to visit
--more ads increase the Adsense program's administrative overhead
--more ads encourage ad blindness (bad for the publisher, bad for Google)
Google's resources (bandwidth, CPUs, storage space, etc.) are not limitless. And the public's tolerance for advertising is not endless. More and more ads impose extra cost, which Google will only want to incur with good reason--i.e., the ads induce clicks, hopefully leading to conversions.
It's hard to imagine Google being completely indifferent to a (huge) site plastered with ads that nobody ever clicks on. Quite apart from Smart Pricing, it's easy to imagine the Adsense system imposing penalties for excessively low CTR.
More ads also means an oversupply.
1) Show ads alternatively to increase chance of click.
2) Show ads where they are visible - above the fold.
3) Use different ad layouts to make sure there's no banner-blindness.
I think the problem is that we do not always have very targeted traffic. Search engine traffic is targeted but return traffic is not and they are not expecting to find something on a particular topic rather they are trying to find topics to read. Which is why it does not make any difference to europeforvisitors - because he always has visitors who are focused on getting very particular info - and the ads are perfectly balanced.
These techniques have always been successful on my sites.
