Hello
Until today I thought that bigger Ad formats tend to do better in CTR and eCPM (RPM) then smaller formats.
So I thought:
- 125×125 < 234×60 < 468×60 < 728×90, and
- 180×150 < 200×200 < 250×250 < 300×250 < 336×280
Of course Ad size depends on the design of a website and therefor a nicely blended smaller Ad can have better effect then an out of place bigger Ad.
- But wait, is it really this linear?
- Are bigger Ad formats really better in earnings?
A few month ago I did a 50:50 split test for about 1 day and compared a large rectangle with the medium rectangle.
In that days I did not have image Ads activated.
The medium rectangle performance was so low that I kicked it right away.
Yesterday, after spending much thought on this I started a new 50:50 split test again with a Large Rectangle comparing to a medium rectangle.
And so far the results are overwhelming.
The medium rectangle CTR is 136% higher then the large rectangle CTR.
eCPM (RPM) is 191% higher.
So first conclusion is that a one day split test is too short.
This is why this one will run at least a week.
Second, not always do larger formats perform better.
The third conclusion is that allowing image&text Ads for a medium rectangle may help it outperform the large rectangle by being a standard sized Ad format which is much more
widely spread then a large rectangle and therefor is in much more demand with more competition per Ad spot.
And fourth it might be that for different seasons (?) specific Ad formats do perform better than in other seasons (a theory to be tested).
What are your thoughts on this?
I have not been using AdSense on my site for a couple of years because I wanted to offer my content in a clean way.
In January of this year I changed my mind and I'm still glad I did.
The first couple of months I opted for smaller ads, mainly 250x250 text ads.
I then decided to try and get the most out of it and added 728x90 text banners which I placed right under my article titles. Those started to outperform the blocks.
I then changed the 250x250 text blocks for 336x280 text blocks. Those showed a much better performance then the 250x250 ones.
I believed to have seen that big ads performed better than smaller formats.
I kept things like this until a few weeks ago.
I switched to text and image ads on both formats, but after a few days switched back to text ads only, mainly because for some reason I didn´t get relevant image ads and the ones that showed performed very poorly.
Yesterday morning I was tired of seeing very low payments since a day or three and decided to switch to a mix of image and text ads for both formats on my best paying pages. I made image only blocks and full horizontal banners and text only blocks and full horizontals and rotated them.
This has made my (modest) income increase a lot, everything skyrocketed, clics, CPC, eCPM went up some 200%, mainly due to those image ads. I'm getting much better image ads than a few weeks ago.
After watching things in this state for some 40 hours I noticed that the 336x280 image only blocks outperform the 336x280 text blocks by large. Those blocks are placed halfway in the article.
However, the same didn´t happen for the image only horizontal image only scrapers at the top of the articles.
For those text only scrapers outperform by a very big margin the image only scrapers.
I have just removed the image scrapers, but I'll continue watching for 5 more days what happens to the rotating text only/image only blocks.
I feel like after those 5 more days I might start a new experiment with different sizes for the top horizontal banners to see if smaller image banners perform better than large image only banners and large text only banners.
One thing I am thinking is that performance of different ad formats varies a lot, what works very well one week might not work at all another. It is good to be checking very regularly and when you see revenue drops make changes, I don't think there's one setup that will always perform well.
My answer is not really very relevant to your message but I felt like it's interesting enough to mention it here.
Btw. Last week I've spent a couple of days creating individual adblocks for every ad I show, combining them with individual channels. I did this after reading your recommendation. It was quite a bit of work but it was well worth it, it makes optimization so much easier, and more complex.:)
Greetings.
Even though in some forums the CTR of a medium rectangle is a little lower than for the large rectangle the CPC prices are way higher and therefor the eCPM is much higher.
I will do this split tests again in January after the Christmas season to see if that behavior changed but until then I think the best way is to go with medium rectangles because a lot more Advertisers using it and therefor the competition is much higher which increases CPC prices.
So if anyone of you is currently using Large Rectangles I strongly suggest you start a split test with a medium rectangle as soon as possible to see if it earns you more.
Read this for more information:
Does the first AdSense Ad Unit really gets the highest paying Ads
All Adsense ads served on my sites are placed on 24 hour hold and then reviewed daily for acceptance/rejection.
This allows me to personally ensure only relevant ads are available in inventory, and also allows me to catch any advertisers who may be pushing 'odd' sized ads not within the primary four.
180x150
300x250
160x600
728x90
And probably the 468x60 but it is not in bold, but it is the ancestor of all ad sizes :-)
I'm also changing several 336x280 to 300x250 on my sites. And will keep those primary sizes in mind.
I would do a split test before completely replacing the 336x280 rectangles.
In rare cases the large rectangle still performs better then the medium one on my websites.