Welcome to WebmasterWorld Guest from 18.104.22.168
I have a big site currently ranked 20,000 in Alexa. Here's what I did.
1. I joined Google Analytics. The program is integrated with Adsense. Adsense gets an accurate reporting about my visitors so I believe it's the reason why it bumps my eCPM.
2. I made changes with my popular pages. I used to put 3 ads in one page. I tried removing 2 and have only one skyscraper ad on the right hand side (left side is even better). For some reason, I get more clicks; my Page CTR went up to 0.58% compared to my usual 0.25%.
3. In my gallery pages, I switched the placement of my ads to the left. They used to be on the right. I notice it's more effective to place them on the left because my users always start looking on the left side when they view the photos/content. Thus, they look on the ads first and if they are interested in the ads, they click on them.
4. Users are much smarter now and they recognize the ads right away. I realize I should not put so many ads and not be too invasive. What I found out is they like great content but less ads. They will click on your ads eventually if they keep coming back to your site.
I completely agree with your suggestions to use Google Analytics [google.com] (and I'd suggest Webmaster Tools [google.com] as well). I'm also all for moving ads to the places on the page where users are most likely to see them. You can check out this heat map [google.com] to get a sense of which placements your users are probably paying the most attention to.
There's one point where I don't quite agree. While reducing the number of ads on your page might have worked for you in this case, it's a move I'd be wary of suggesting to others. We've found that publishers overwhelmingly do better with three ads than with one. There are always exceptions, but reducing the number of ads on your page is not something I'd personally recommend.
Seriously, congratulations on seeing your hard work pay off.
Anybody else have success stories or tips to share?
Not to highjack your thread, but, I must be one of those exceptions ASA referred to. As soon as I try to add more adblocks my CPC for those blocks plummets.
I'd go as far to say that my overall cpc eventually drops also.
I used to run a site with about 25 or more adblocks and when gradually, i cut them down my cpc and my total income increased. Over time I removed most of the remaining ads whenever faced with minor downturns and came to a point where a couple of ads on my best performing pages achieved reasonable stability for a couple of years.
The current downturn adds a new dimension, however.
I'm not advocating that what I did will work for everyone
[edited by: Scurramunga at 2:32 am (utc) on Mar. 21, 2009]
We made some changes about a week ago with the new custom search. The new search box a well as a change in placement has resulted in a four fold increase in revenue (although this is very small compared to other ad forms).
That would relate to the number of ads shown. I think you’re losing on that end. If it was no change, or some improvement the number should be .75 or higher. Those extra ad blocks lower on the page were bringing in lower paying ads so you’re not seeing those now. That would explain the higher ECPM. Correct?
My theory is that when there are text ads - the ads that are higher on the page pay more (I believe ads are displayed in order of the advertisers max CPC right?). When there are more ads on the page, there are more low paying ads - so even though u might get more clicks when more ads are on the page, they are often lower paying.
When I have less ads on a page, my average CPM is much higher, but my CTR is lower - overall I'm making more.
I mean no offense here ASA - but I hear your message quite often on adsense blogs about placing more ads but I constantly see in forums people saying less ads is performing better. I wonder if it is just google policy to promote people to place more ads ;) I can imagine that would be a good thing to get more advertiser impressions.
I joined Google Analytics. The program is integrated with Adsense. Adsense gets an accurate reporting about my visitors so I believe it's the reason why it bumps my eCPM.
I find that hard to believe. If that were the case, we would have heard this from other people. I think the reduction in the # of adblocks is sufficient explanation.
Are you also suggesting that we remove the content too and disable the back button - I hear those trick greatly increase ctr. No we can't do that since that would be the same as domain parking.
joined:July 3, 2008
All ASA is pointing out is that webmasters need to experiment and figure out what works on their sites. I run three ads on each page but the pages look clean because the ads are relatively small, but well placed.
The average page has about 1,200 words on it, so each page is very long and if anything the pages look pretty empty of advertisements.
Pretty soon they'll be encouraging arbitrage again :)
At the top half of my page I have the navigation on the outsides and the content between them(free software download site). On the bottom half of the page(entire page is above the fold) I have 3 blended 250x300 adblocks set to image and text. I thought the same thing less maybe more and put a 728x90 in their place, a total disaster! The downloads average 2 minutes and the general user only starts one download at a time. Nothing else to do but read the ads that are specifically targeted to the page content. b2b> b2b sports>sports etc 10 pages, any ad that does not fall under the subject matter is blocked (lots of work) The users were actually disappointed that there wasn't enough ads for them to read through and pick one, imagine that!
Pre recession this was a multiple 6 figure site with evergreen content now it's down to mid five figures. Hanging in there just very a painful position right now. (I do this full time and yes I have 10 more sites but that is the breadwinner)
Also the Advertiser money is there but my CTR is 30% of what it used to be. People just don't have the money so they don't click but the still take the free software, just going to have to wait it out.
[edited by: Khensu at 5:48 pm (utc) on Mar. 22, 2009]
Anybody else have success stories or tips to share?
ASA, I would share my story if I considered it a "real" success.
Earnings are good for me but what's stopping me from calling it a "success" is the complete lack of any certainty regarding the income due to a specific factor: smart pricing.
I understand it might be a useful tool but the fact the publishers have no kind of knowledge or control about it and still are penalized for it causes too much "instability" in the system.
An ad that doesn't convert well might have a million causes, among them:
- Google delivered an ad that wasn't exactly in theme with the content
- The ad was misleading and the landing page showed a product the user didn't expect
- The landing page itself is badly designed and doesn't help conversion
... and so on. But it is the publisher that get penalized and no one else.
I could add I find unfair that it is the entire Adsense account that get smart priced, including sites without low conversion.
Also, if the ultimate goal of an ad should be to convert, why doesn't google give us more information so we can take the appropriate actions about it and improve it?
I mean no offense here ASA - but I hear your message quite often on adsense blogs about placing more ads but I constantly see in forums people saying less ads is performing better. I wonder if it is just google policy to promote people to place more ads I can imagine that would be a good thing to get more advertiser impressions.
Ya, I figured out what is acceptable about 4 years ago - one ad mostly and occasionally two on larger pages. I just felt that originally Google encouraged smothering the visual hot spot with ads and then they backed away from that recommendation and it is funny to hear ASA encourage it again now.
I would like to perform a similar experiment on websites I run, but downloading, amending and uploading files from a large scale website which does not have a CMS can be a bit of a chore.
Can you tell me if I would be violating the AdSense TOS if I were to place a "display:hidden" command into my CSS in order to switch on and off the AdSense units? Each of my AdSense units has it's own class or ID attached and being able to control the units via CSS would save me a hell of a time and allow me to perform many more experiments with AdSense.
Thanks in advance.
I've tried the "less is more" approach.
I've tried the "more is best" approach.
Result: For any given site on any given day, I wouldn't bet the farm on whether "less" or
"more" is better.
If I had an extra hour today to spend on testing the "less" approach vs. the "more" approach, I'd decline and instead spend the hour making one of my sites better for my visitors.
Back to work.
I was hoping for some real meat-and-potatoes kind of thing where they would refer you to the heat map, explain positioning, give you some examples of other sites that they knew about that implemented Adsense well. Instead all I ever got was just paste extra Adsense units on your pages. I never once saw any evidence that doing this marginally or substantially improved income for my site. It just made my site look more spammy (oh yeah, they were big on using image ads too).
The last time they offered to let me participate in this program I just ignored the invitation.
It's been a month since I first posted this thread so I'd like to give you an update.
So far, my earning is steady and doing very very well. I expected it to go down because of the economy but surprisingly, my daily earning is up, sometimes way up.
I can't pinpoint exactly what I am doing right. Perhaps I am getting many advertisers because of great content? Or because of the ad placement changes that I made? Or the lesser number of ads on a page? Or Google Analytics? I honestly don't know.
All I can say is that the changes that I made seem to be working so I'm sticking to it and not make any more changes.
But I'm not advocating that what I did will work for everyone. Just experiment and experiment and see what works for your site.