|Result of adding a second ad unit.|
Here's what happens when you add a second Google ad, Clickthrough rates crash. I went from about .5 to .2, and my Effective CPM has also been cut more than half. Thats because each ad counts as one page impression and of course the bottom ads will be clicked way less than the top ads. Each page view generates two page impressions. If you show three ads per page it'll crash even more!
If Google terminates affiliates because of CTR or CPM rates, there might be a lot of us that end up geting that dreadful E-mail.
Before this change I averaged 15,000 page impressions a day, and with nine hours left today, I've got 39,056 'Page impressions', er ad impressions. By the end of the day it'll probably be more like 70,000-80,000 with about 30,000 of that being from adding the ads to my message boards. Still a huge increase from 15,000!
I think they should change 'Page impressions' to 'Ad impressions'. Earnings wise it looks like it's staying about the same. I'll need a few extra days to see how much of the earnings come from the channels for the second ads, and message boards.
This also increases the number of PSA ads, or your own ads, BIG time. Before this I was geting my own ads to show up about 1,000 times per day, and today it looks to be around 10,000 ad spots showing my own ads. I don't think those numbers are included in the page impression number, so it might get close to 80-90,000 total 'page impressions' counting my own text ads.
My revenue doubled on the 2 pages I've added a second ad unit.
I didn't add any ad unit on the other pages, no space for them. But I like it.
|Here's what happens when you add a second Google ad, Clickthrough rates crash. |
Interesting to start seeing some feeback. Having the CTR go down doesn't surprise me and really isn't the important stat. What is important is the impact on your income. Did it go up?
|If Google terminates affiliates because of CTR or CPM rates, there might be a lot of us that end up geting that dreadful E-mail. |
I'm sure Google has already considered that CTR and CPM would drop. It shouldn't be a big deal.
Here's my stats:
Impressions up about half (even though I've tripled my ad panels)
Number of clicks about the same or marginally higher
CTR down 30%
EPC down 32%
EPM down 55%
I would say my overall revenue has *dropped*, despite tripling the ad space. This is mostly due to the drop in EPC. Even if revenue has marginally increased, it's a lot of trouble to run all those ads.
More importantly, I suspect that there is an "EPC penalty" for low-CTR sites, and running excessive ads would cause a low CTR.
I'll probably switch back to one ad panel over the weekend.
:::Earnings wise it looks like it's staying about the same.
Though now it looks like it'll be atleast a few bucks more a day. In about five hours and also when channels are updated, I'll have a much better answer. With five hours left in the day, the last day that I had as many clicks as today so far was almost a month ago. So it looks like this will increase earnings (Some of the increase is probably from the message boards actually generating clicks!)
Here's an E-mail I just sent them! Had fun with the Gooooooooogle!
On the reports page can you change 'Page impressions' to 'Ad Impressions'? With two ads on the page it doubles the number even though page views stay the same.
Abbbout thhhhhe Gooooooooooooooooogle, yoooooooou miiiiiight wannnnnnnt toooooooo reeeeeead sooooooooome webmaaaaaaaster booooooooards. Moooooooooost webbbbbbmasterrrrrrs dooooooooooon't liiiiiiiiike it.
As long as it doesn't decrease earnings, it doesn't bother me. :)
From the AdSense glossary:
"Page impressions are used when calculating reports for AdSense publishers, rather than ad impressions."
And from the AdSense FAQ:
An ad impression "is generally the equivalent to a page view."
Does Google count one page impression each time a page with AdSense ads is loaded, regardless of how many ad units are on the page? Or will a page with three ad units now count as three impressions each time the page is viewed? My numbers would indicate the former -- the number of impressions has stayed roughly the same for pages with multiple ad units.
|Does Google count one page impression each time a page with AdSense ads is loaded, regardless of how many ad units are on the page? |
From Google's Glossary
Ad unit - A set of ads displayed as a result of one piece of the AdSense ad code.
Page Impression - A page impression is counted each time an ad unit is displayed on a publisher's site.
Sounds like a page impression is counted for each piece of Adsense ad code on your page. If you put three ads on a page, three page impressions are going to be counted. Just my speculation based on Googles' own information and the experience of Jesse_Smith and others.
Yes, but. Maybe they just need to update their definitions? I can't see why they'd want to count each ad unit viewed as an impression -- G has never made any distinction between ad units that show four ads vs. ad units that show one ad when it comes to counting impressions.
Well, for the full day the number of clicks I got was almost double the average, but earnings were about avearge! Total 'Page impressions' were 79,346 compaired to about 15,000 a day before. My PSA replacement ads showed up 11,163 times compaired to about 1,200 before.
CTR crashed to .2%. Before that it was between .4 and .7%. CPM also crashed, BIG time! Let's say it used to be $1.00. Yesterday would of been about $0.20. Hopefully half the earnings havn't been added yet!
I see the same ad twice on one page. Exact same title, desc, and url display.
Just to give my feedback:
I had run a wide skyscraper on the top right side of one my content-rich pages that gets a high click-through rate, and a high EPC. I had used a table to create a second column for the skyscraper.
There is a lot of content on this one page, and I know visitors make it to the bottom as I have links there that get clicks at times. So, I figured this would a great opportunity - I went and added a second wide horizontal banner in the middle of the page, and another at the bottom.
First thing I noticed was that the skyscraper was showing defaults, but the first horizontal ad in the middle of the page had ads. I did a little reading and noticed that the first ad to show in the HTML code was the one that would get the first ads.
So, I decided I would get rid of the wide skyscraper and replace it with a horizontal banner at the top.
I was able to check the channel data today from 9/2/04.
Basically, the horizontal ad at the top got around 1/2 the number of clicks that the wide skyscraper ad had been getting. More interesting, was that the second and third ad units on my page got NO clicks.
I figured that the top ad probably got fewer clicks because the skyscraper just attracted more attention. So, given that I have a long page, I figured I'd put three skyscrapers on my site - one on top of another. I was curious how this would look, figuring it would be nice to have ads going all the way down the right side.
So, the first ad unit shows a default ad, and the second and third unit show nothing. I've since removed the second and third unit, figuring it was messing something up, but now my first ad is showing ads half the time and the default ad the other half.
I don't know if something I did with messing around with the ad placements has caused adsense to decide not to run ads on that page - hopefully not.
For now, I think I'm going to go back to having the code for 3 ad unit skyscrapers down the right side of the page and wait for another few days to see how the data look.
3 days of adding total of 3 ads now on over 2000 pages
CTR down from 3% to 1.4% (of course this will drop when adding more ads)
EPC around the same
Earnings up 35%
all in all, this is better for us
Just sent ya a sticky.
:::Hopefully half the earnings havn't been added yet!
I think I was right there! Today I've got half the clicks I got yesterday, yet my earnings are allready 20% higher than yesterday, even with half a day left! Hopefully earnings take an extra day to show up! If each day ends up like today looks like it'll be, I'll be really happy!
I don't care what the naysayers say, this stuff is really fun! And super educational!
So many excellent comments, and data points.
On one page where I put a second ad panel, I got GREAT results. Google gave me the always-presented seemingly unavoidable lousy ads on panel #1, which got them out-of-the-way, so to speak, and the ads I have been wanting FINALLY appear in panel #2. If you have a similar problem, give a second panel a try. My latest mini-theory is that sometimes you gotta move past the "always-presented-crappy ads", to get to the gold. Anyone seeing this at all? I think 3 ad panels is over the top, until you see just what 2 ad panels will do for you!
<edit>I just saw a great presentation of 3 ad panels on a site, that look really great - leaderboard at the top, skyscraper in the middle(right side), and another leaderboard at the bottom. Broad generalizations (especially mine) must always be questioned. I'm a "conclusion jumper". As always, it seems, "it" depends. On many things. Ad inventory, keyword specificity, Google adsense algo, site design, business area, ad business dollar level, competition, super-b¦tchin-algo-change-whoops-put-it-back-real-quick, etc.
That's what makes it so much fun. Kind of an online IQ test. Ha! Let's play!
What should I do next? Hoooooooo, I don't know. But WebmasterWorld will tell me! Just keep reading, and keep an open mind.<edit>
I have to reconsider, too. At first I wasn't thrilled but now that channel stats are in I can see that the ads I did not have a week ago are the ones being clicked on. This will, at the very least, be a great way to find out which location and style is best for an ad becuase you can experiment with different options on a single page and see what's clicked most.
I went back and looked and my left column skyscrapers were the ones not being clicked. No, it was the new ones waaayyyy down at the bottom and in the middle.
Earnings are in the toilet as is my CTR and CPM but the new ads are th eones being clicked. Hopefully this is just temporary and completely normal and everything will be back to normal after the holiday.
Over the past year people have been bemoaning the fact that they make almost all of their money from Adsense and that was too risky. Now that Google is allowing three ad units per page why would you jump on this? If you think Goooooogle looks childish, what do you think 3 Adsense banners on one page looks like? Pathetic and unprofessional. I'd rather stick with my one adsense banner and sprinkle in affiliate links here and there to balance out my risk. And guess what? My EPC, CPM etc. for Adsense has not stayed the same the last few days, it's gone up...Just because something new comes out doesn't mean you have to jump on it and implement it. If it doesn't work then remove the additional adsense banners from your pages...easy solution!
your max CTR on pages with 2 units (if a user only clicks on ad per load) would be 50%. With three units the max would be 33.3%. CTR should certainly be cut in half on pages that you have 2 units on.
I have seen the lowest earnings since joining adsense today, unless it picks up real real quickly in the last few hours of the day. I mean my earnings are not even close to regular. I am about 10% of regular daily revenue.
I think this is a poor time to compare day to day figures, given the Labor Day weekend.
IMO, double or triple ad units may not be a good idea. I did put a 2nd unit on some pages but stopped doing more after thinking about it some more.
Keep in mind the first unit on the page (from top to bottom) would seem to contain the best paying ads. The 2nd units likely pay less, and 3rd even less with the lowest epc adwords, assuming the same category/keyword targeting for all units.
So if you get the same number of clicks per page with multiple units chances are your income will decline if the clicks are now split between high paying ads in unit 1 and lower paying ads in unit 2 or 3.
Just my opinion on this as do not have evidence it is so however it does make sense to be correct, right? A big question is if multiple units will result in more clicks per page, or not? Any opinions here on that issue?
Trader I'm seeing exactly what you described. Same number of clicks but my bottom line is down 50%.
Keep in mind that it looks like Google has had some problems displaying ads (maybe due to the succes of the 1-to-3-ads-per-page transition). This could also have had an effect on the results. I can imagine that results will increase the coming days, since it looks like the capacity-problem had been solved.
Well, I had the channel data from 9/3 to look at now. It appears the click through rate for a horizontal extra wide banner at the top of my page got just over half the CTR as a wide skyscraper on the top right side (I assume giving general "relative" figures about my data isn't against Google's TOS, as I'm not giving any specific numbers, and if anything, this information should help both users and Google - if it is, hopefully someone will warn me).
The wide horizatonal in the middle of a long content page surprisingly had about 10% the CTR as the horizontal banner at the top of the page. I had heard other people mention that adding the banners more into the middle of the content will often increase CTR - perhaps I just chose the wrong spot in the middle of the page to add the banner.
The wide horizontal at the bottom of the page had about 20% the CTR as the horizontal banner at the top of the page.
So, yesterday afternoon, I got rid of the horizontal banners and replaced them with 3 wide skyscrapers down the right side of the page (it is a very long page) - with them being spaced out so that as one banner is scrolling off the page, the next one shows up.
Today, it appears Google is sending me default ads about half the time, and the other half fills up all 3 wide skyscrapers (15 ads). Hopefully this is a problem on Google's end with the defaults - but I'm only noticing defaults on a small minority of my pages. Perhaps it is just Google adjusting to the multiple ad units on the page.
While the data from the holiday weekend may not be the most applicable to estimate future performance, at least I'm not losing too much if things are broken, given the holiday weekend's low traffic overall.