| This 63 message thread spans 3 pages: 63 (  2 3 ) > > || |
|Remove AdSense from pages with low eCPM|
take out those ads with eCPM < 1 and you might recover from smart pricing
My experiments may not be conclusive.
I have sites/pages with very different eCPMs. Earnings used to be ok until smart pricing hit hard on me recently. With not much to lose, I decided to do something about it. One way, as suggested by some, is to remove AdSense from pages that generate low earnings, but how do we identify such pages? A page may have high EPC (therefore high eCPM), but because of low impressions, it generates low earnings. Other pages may have low eCPM, but give decent income. I went with low eCPMs.
After removing AdSense from pages with extremely low eCPMs (< $1), overall earnings started to rise. But when I restored AdSense on these pages, "smart pricing" hit again. Seems to me that Google hates low eCPMs, so much so that it smart-prices entire accounts of publishers.
Does anyone have similar experiences?
Yeah, I'm considering removing ads from an entire site(10K+ page view a day) and replacing the ad unit space with ads for my own higher performing pages(on a different site). Either that or go with a more aggressive Adsense unit to improve CTR. I haven't decided yet. eCPM is a little over $1.00 but not much. It went down recently due to a change I made, I thought it would help, but instead it just made things worse. I guess, as they say, just try it and see what works for you.
Yes, and I have seen several others report the same. Though I looked at CTR, the pages where I removed the ads were most likely also very low eCPM.
There's been a lot of debate about what factors Google considers in smart pricing in addition to conversions (or instead of conversions, when that information isn't available) and MOST people think that Google couldn't possibly consider CTR or something so strongly linked to CTR as eCPM is, but I'm not so sure.
Certainly it's NOT a strong effect, so that you can hurt your earnings if you pull ads from too many pages, and it also may only work on some kinds of sites.
But if it works for you, go for it.
I haven't seen any benefit in removing AdSense from low-eCPM pages, but--as the expression goes--your mileage may vary. Smart pricing is probably like Google's search algorithm: Multiple factors are taken into account, and something that has an effect on one publisher may not hurt--or help--another. Fortunately, it's easy enough to test (although this month's test results may not predict smart-pricing behavior a month from now).
Oh, and if you are getting any advertisers who are targeting your site with CPM ads, then removing ads could be counter-productive. My site has yet to be targeted. If it was, I would experiment with putting ads back on certain pages.
|There's been a lot of debate about what factors Google considers in smart pricing in addition to conversions (or instead of conversions, when that information isn't available) and MOST people think that Google couldn't possibly consider CTR or something so strongly linked to CTR as eCPM is, but I'm not so sure. |
I think the official word from google is that ctr is NOT a factor in smartpricing, and when I'm in doubt, I usually believe google.
It makes sense. HOWEVER, CTR may not have an impact on smartpricing, but it could very well have an impact on AD SERVING, for better or worse. It could factor in to WHAT ads are served, but frankly, I have better things to do than speculate and read the algo mind.
Rather than tweaking placements all the time, I've decided to place ads (and number of ads) according to a) the design of the specific site and b) user comfort levels. If there's some sense in having an ad block at the bottoms of long articles, I do so. On the other hand it doesn't make sense to add an ad block at the bottom of a short page, since it makes the page look MFA, and I don't want my visitors thinking my pages have more ads than content.
>>>A page may have high EPC (therefore high eCPM)<<<
no, high epc does not always mean high ecpm... but you will typically see low ecpm with low ctr... that's why ecpm is not very useful for troubleshooting, or much of anything else.
if you are going to do something, pull adsense blocks, or remove adsense entirely, off of the pages with low ctr first... that's what worked for me.
|that's why ecpm is not very useful for troubleshooting, or much of anything else. |
eCPM/CPM is an industry-standard way to measure revenue from pages and to compare revenue sources. It's extremely useful for those purposes, since it helps you determine (for example) whether the space devoted to an AdSense ad block or a display banner could be used more productively in other ways.
|if you are going to do something, pull adsense blocks, or remove adsense entirely, off of the pages with low ctr first... that's what worked for me. |
And just for you (I've been meaning to mention this) make sure that your site is coded well enough so it doesn't absolutely break in firefox, so badly that the result is that pages violate the adsense tos about obscuring adsense ads.
It's a neat example of how someone can be disappointed in adsense and google, but have things on his or her site that are pretty much guaranteed to result in lost income.
You should check it. It's on one of the vid. sites.
I'd want those firefox people to be properly served, and I'd be more concerned with that than pulling an ad or two here and there.
>>>eCPM/CPM is an industry-standard way to measure revenue from pages<<<
adsense does not base its payout on ecpm/cpm rates, so it's not relevant.
unless your page values have sunk so low that adsense cpm is all that you can get, that is.
Let me use a Publishing 101 example to show why eCPM/CPM is an industry-standard metric that works with any kind of ad (CPC, CPM, or even a flat monthly sponsorship rate):
- You've got a review site about widgets that has three types of ads on most pages: an AdSense skyscraper, a FastClick leaderboard, and a Widgetrex affiliate ad.
- You do the math, and you see that your AdSense ads are earning an average of $X per 1,000 impressions, your FastClick ads are earning an average of $Y per 1,000 impressions, and your Widgetrex affiliate program is paying an average of $Z per 1,000 impressions. Immediately, you know which ads are performing best, and which ones might be worth replacing (at least on a test basis) with alternatives. Why? Because you're comparing apples and apples (eCPM/CPM) rather than apples and oranges (EPC vs. CPM, for example).
eCPM can also be used to compare performance of ad units (such as Google AdSense ad units). Google's "Inside AdSense" explains this at:
not a flame, but the the o.p. of this thread already proved my point: "A page may have high EPC (therefore high eCPM)" which is not a true statement... people are much better off ignoring ecpm when it comes to adsense.
mixing cpm banners with adsense is a good way to run your adsense epc right into the ground... for example, efv, that's how you do it, and didn't you just tell us that your epc was down by 30%?
do you have any pages where there are no adsense blocks?
if adsense won't work on a page when it's by itself, mixing it with a banner can only make things worse.
you don't need ecpm to know when to pull adsense off of the page completely, because your ctr and epc channel data already told you exactly where the problem is... ecpm can't do that... same thing when comparing adsense blocks on a page.
Danimal, you're welcome to believe anything you like, but some of us come here to gain and share professional knowledge.
Fact is, eCPM/CPM is a standard industry metric, and it's a very useful one. It allows comparisons of revenue streams from CPC ads, CPM ads, affiliate programs, flat-rate monthly fees, and other types of pricing. As Google explains, it can also be used to compare earnings from different types of AdSense ad units. Big publishers use eCPM/CPM to measure ad performance, and there's no reason why small and medium-sized publishers shouldn't learn how to benefit from using it, too.
Addendum: eCPM can also help in setting rates for ads that you sell direct. By looking at the eCPM in your AdSense channels, you can get an idea of what traffic is worth on different areas or pages of your site. This makes it less likely that you'll undercharge for placement on high-value pages or risk losing a prospect with unrealistically high rates on low-value pages.
>>>some of us come here to gain and share professional knowledge.<<<
telling forum members to do things that could drive their epc down by 30% is not professional knowledge.
>>>It allows comparisons of revenue streams from CPC ads, CPM ads, affiliate programs, flat-rate monthly fees<<<
the o.p. never asked about CPM ads, affiliate programs, or flat-rate monthly fees... the topic of this thread is removing adsense from pages that don't perform, and ecpm is the wrong tool for doing that.
|... the topic of this thread is removing adsense from pages that don't perform, and ecpm is the wrong tool for doing that. |
On the contrary: eCPM is the tool for measuring a page's performance in comparison with other pages. But I guess we'll just have to agree to disagree on that and let other publishers draw their own conclusions.
Danimal, it's true that EPC and CTR give useful specific information, but I think that also eCPM is a very useful metric of general ad performance, as EFV explains.
We can remember the basic equations:
Earnings = eCPM * Impr / 1000
Earnings = CTR * EPC * Impr / 100
eCPM = CTR * EPC * 10
That is to say:
From factors such as ad relevance, placement, topic-focused pages, ad blindness, etc. ===> CTR
From factors such as advertiser conversions, bids, niche, smart pricing, etc. ===> EPC
From CTR and EPC (specific metrics) ===> eCPM (a general metric)
From eCPM and impressions (traffic) ===> Total earnings
Apart from this, also in my case, I've seen -almost always- increased total earnings after removing ads from very low eCPM pages. Usually, some days or one or two weeks after the removal. Google seems to be happy with sites who avoid very low eCPM.
juan, that's a great post, but ignoring specific info like ctr and epc means that you ignore an opportunity to fix what is wrong with the page, and even how it affects the r.o.i. of the entire site or account.
and i'm not flaming the o.p., but we have already seen how ecpm confuses people... in both this thread and many other threads.
i'm glad that ecpm-only worked for you, but the efv website is the poster child for what can happen when you base decisions on ecpm.
|i'm glad that ecpm-only worked for you, but the efv website is the poster child for what can happen when you base decisions on ecpm. |
LOL. That's pretty funny, coming from a member who has complained repeatedly about single-digit EPCs.
I'd say more, but every thread in which you resort to ad hominem attacks seems to get deleted these days, so it isn't worth wasting my time.
|That's pretty funny, coming from a member who has complained repeatedly about single-digit EPCs |
Sites that break in the second most common browser, so that the ads get obscured, and the site breaches tos are going to have those kinds of problems, I'd suspect.
It's easier for some to complain about google than to fix their own sites, or realize their MFA look-alike contentless sites don't provide a model to get anything more than single-digit EPC.
I'm dreaming of the day when inferior, deceptive sites will only be able to advertise on other inferior, deceptive sites, and all the ads they will get to display are from other inferior, deceptive sites paying a penny.
I know it's a dream, but from looking at those inferior deceptive straight MFA sites, it sometimes looks like the ads they are getting are from other similar sites. On occasion it looks quite strange.
>>>LOL. That's pretty funny, coming from a member who has complained repeatedly about single-digit EPCs.<<<
i don't recall that, but i can tell you that my september epc was the highest it's been all year... maybe even the highest that it's been since i started with adsense.
so there are 4 of us in this thread so far who have had success removing adsense off of pages that don't perform, you should probably listen to our advice.
|so there are 4 of us in this thread so far who have had success removing adsense off of pages that don't perform, you should probably listen to our advice. |
That might be valid advice for sites that don't attract site-targeted CPM ads. However, if you've got a site that advertisers want to be on--e.g., a site that doesn't fit rbacal's description--that may be a shortsighted approach, because it limits potential revenue.
|That might be valid advice for sites that don't attract site-targeted CPM ads. However, if you've got a site that advertisers want to be on--e.g., a site that doesn't fit rbacal's description--that may be a shortsighted approach, because it limits potential revenue. |
To add. If, like danimal your ecpm is dreadfully low, then site-targeted ads will also probably pay really really dreadfully low. So why keep non paying blocks?
On the other hand, let's say I have a site that has a $18 CPM that has regularly hit the $32 CPM range. For site targeted ads, the cost to advertiser is going to be way high (and certainly worthwhile to me) even given that it might appear in a lower cost ad block. In this instance, it's advantageous to have the extra adblock there, particularly since google has said officially and publicly that ctr is NOT a factor in smart pricing.
Different solutions for different situations. What works for a borderline sleaze site would be stupid on a content rich, reputable site, and probably vice-versa.
|On the contrary: eCPM is the tool for measuring a page's performance in comparison with other pages. But I guess we'll just have to agree to disagree on that and let other publishers draw their own conclusions. |
I'd say it's a tool for measuring ad performance, BUT when considering if you want to remove ads from a page I would strongly suggest that you do not consider it in isolation. You need to look at ALL metrics (including logfiles) to accurately asses if ads work or not.
The problem I have with ecpm is that it frequently gives wild numbers that do not represent the true picture. For example, some days I may see ecpm of over $1000 on a page. That may indicate a page is a good page for ads, or it may represent a complete abberation. To work out which, you need to look at traffic, clicks, epc where traffic comes from and quite a few other factors. On it's own it's not that useful.
ecpm is a useful measurent, but without the context of other metrics as well it's not of much use in making the decision to remove ads or not.
>>>That might be valid advice for sites that don't attract site-targeted CPM ads.<<<
you keep telling us how good site-targeted cpm ads are, but then you admit that your epc has tanked by 30% since last year.
it looks to me like your adsense blocks are worth a lot less than they were last year(30% less?), so of course the cpm can win some bids to get in there... bottom line, it means a lot less profit per ad block for you.
so yeah, you are correct efv, i bet that's "attracting" all kinds of advertisers ;-)
I have found that I do well by removing pages that I have attract low paying clicks. Ecpm is usually a fair indicator, however be cautious before drawing any conclusions.
Danimal, I've made it clear in other threads that the moderate decline in my AdSense EPC for September and October is likely to be the result of changes that were made to maximize overall revenues along with a substantial increase in pages (and therefore AdSense clicks) on relatively non-commercial topics. Think about it: If a site adds 500 pages on Elbonian walrus-watching while increasing the likelihood that users will click on affiliate links instead of AdSense ads for more profitable topics, its AdSense EPC will and should decline. You may not understand the concept of fine-tuning the revenue mix to maximize overall earnings, but that's to be expected since your sites--or at least the ones I've seen--rely solely on AdSense.
Scurramunga, you may have a good point about pages with low EPC, but the original poster was talking about pages with low eCPM (which can be the result of low clickthrough rates, not just low EPC). On my photo galleries, for example, I tend to have low eCPMs from CPC ads because users tend to look at a lot of photos without clicking (or at a lot of photos before clicking), but the EPC is just fine.
For what it's worth, I've experimented with and without AdSense ads on my low-eCPM (but decent-EPC) photo galleries over the past couple of years, and I've never been able to discern any benefit from pulling the code. As with most things, different people may have different experiences, so testing is the best way to find out whether pulling AdSense code from low-eCPM pages will improve or hurt total AdSense revenues.
I have found that I do well by removing pages that I have attract low paying clicks
If you remove low paying ads blocks, you increase the probability that people may click on high paying ad blocks, even on other pages, which certainly can account for higher ctr on those remaining ads, and thus higher CPM. The higher the CPC on an ad block you want to remove the greater the effect will be.
Whether that happens on a specific site can be examined empirically. Sorta.
Again, unless people believe that google outright lies, CTR has nothing to do with smartpricing.
>>>Danimal, I've made it clear in other threads that the moderate decline in my AdSense EPC for September and October<<<
your said that your epc dropped by 30% from last year, which is not a "moderate decline", it's a disaster.
>>>If a site adds 500 pages on Elbonian walrus-watching while increasing the likelihood that users will click on affiliate links instead of AdSense ads for more profitable topics, its AdSense EPC will and should decline.<<<
no, the overall adsense ctr would go down, but the average epc can only decline if you put up a bunch of pages that don't work well with adsense.
if your channels were set up properly, you should be able to see how much the new-page epc is driving down the average epc for the old pages.
They dynamics (I'm being polite here) that drives people to try to win their arguments by hacking at each other with what amounts to butter-knife edged thinking really ruins the use of threads that start out great, and then get hijacked by the same folk saying the same things.
...is there any chance that we can stop the BS divergences, and talk about the TOPIC?
If we stipulate, danimal, that you are right about everything, will that help you out?
Will you stop trying to assassinate (using your butter-knife) EFV, and a t least let others who want to talk about this topic proceed in a non-personally aggressive manner?
| This 63 message thread spans 3 pages: 63 (  2 3 ) > > |