Posted: Sun Aug 17, 2003 5:36 am Post subject: AdSense colours
Now that Google allows you to customize the colours of the AdSense banners: has anybody noticed differences in performance from differently coloured banners?
I'm in the middle of a test run right now, and the best one so far has actually performed more than twice as well as the worst one (unfortunately the best performer is the ghastly pink scheme called "It's a girl" - I really hope one of the yet untested colour schemes will beat it )
The test is by no means conclusive, though. I only run each colour for one day. _________________ Frank Nordberg
http://www.musicaviva.com
The internet center for free sheet music
Last edited by Frank Nordberg on Tue Feb 03, 2004 4:09 am; edited 1 time in total
There have been a lot of posts on this subject, here and at other forums. There is really no one color or size that will benefit all sites in all topics.
I want to experiment on my site over the next few months with sizes and colors. While I can't differentiate Adsense stats to tell where the clicks are coming from, I am trying to increase the overall CTR.
What I propose to do is use PPC testing to see what works best (well as best as I can determine with limited stats). I plan to take new pages (noindex,nofollow), write a simple ad, and run PPC for one week with Adsense at x size and x color, another week at y size and x color, etc. and track changes.
I agree it isn't perfect, but felt it would give me a start on testing something like this.
Any thoughts or suggestions to improve my testing?
Debs _________________ Learn how to turn keyphrases into quality, well-targeted articles your visitors and SE's will love with Gary Antosh's new ebook "Web Content Made Easy!"
Now 2 month after Frank's posting, I wonder if anybody else has experienced improvements when changing color's on ads.
Somebody else (sorry don't remember who) mentioned a similar test in another thread, and the topic's been mentioned in various other circumstances here too.
After two test runs the conclusions when it comes Musica Viva is quite clear:
Colour does matter. The best ones performed more than twice as well as the worst ones.
The best performers were the ones that matched the overall colour scheme of the page well without blending completely in. "Discreet but visible" seems to be the rule.
Since Musica Viva uses very light greytones this means light, primary colours with no borders. Judging by comments by others here and what Google use themselves, this seems to be a general trend, although not necessarily a universal law.
Cycling through colours for more variety does not seem to improve the results. In fact the colour I ended up with (Aquadoodle) performs even better now than it did during the test.
Excellent Frank and thank you I looked over your site and will be taking your testing under consideration for my own site.
Debs _________________ Learn how to turn keyphrases into quality, well-targeted articles your visitors and SE's will love with Gary Antosh's new ebook "Web Content Made Easy!"
Any thoughts or suggestions to improve my testing?
Just don't be too quick to jump to conclusions based on small sample sizes (<1000). This is the age old problem, when you're trying to test lots of small changes. It can be impractical to test anything other than major changes via the PPCs for reasons of cost and traffic levels being too slow.
I am fascinated by the psychology of colour, too.
I have used a lot of blue (for reassurance), with some red as second colour for text that is especially important. Odd words, nothing more - the rest stays black.
I have also tried pastel shades (as opposed to white) as background colours, but modern monitors don't glare as much anyway.
I am also planning to test orange (optimism) as a main colour, but my new sunglasses haven't arrived yet.
All the best,
Charlie. _________________ "Before I speak, I have something important to say."
- Groucho Marx
Charlie: I am also planning to test orange (optimism) as a main colour, but my new sunglasses haven't arrived yet.
Funny you should mention orange, that's the color (however very pale ... I call it orange cream) that I am considering on my site. Hopefully your sunglasses will be here by the time I get it online
Debs _________________ Learn how to turn keyphrases into quality, well-targeted articles your visitors and SE's will love with Gary Antosh's new ebook "Web Content Made Easy!"
Just don't be too quick to jump to conclusions based on small sample sizes (<1000). This is the age old problem, when you're trying to test lots of small changes. It can be impractical to test anything other than major changes via the PPCs for reasons of cost and traffic levels being too slow.
Very good points Charlie. I was in a fortunate position to do these colour tests. With around 20000 daily page visits I could get statistically relevant numbers pretty fast and since I have a "regular" job I love, risking the ads income from my site for a period didn't bother me that much.
Since you can't get page specific stats from AdSense, the results from a test that isn't site-wide will be very difficult to interpret.
Quote:
I am fascinated by the psychology of colour, too.
Hmm, that could actually be a good suject for a site. Judging by the ads Google keeps serving my html colour table at http://www.website-helper.com/handbook/html-color-chart.html it might be a lucrative niche too. (Btw. any ideas how to get AdSense to serve relevant ads to that particular page is higly appreciated )
Quote:
I have used a lot of blue (for reassurance), with some red as second colour for text that is especially important. Odd words, nothing more - the rest stays black.
I ought to mention that my ad colour tests didn't show any difference between various basic colours. Red, green, blue, grey, yellow all worked just as well. The effect seemed to be solely dependent on visibility. Too similar to the background and people didn't notice the ads, too different and the "banner ad blindness" set in.
Quote:
I have also tried pastel shades (as opposed to white) as background colours, but modern monitors don't glare as much anyway.
That is true, but even so it may be a good idea to tone the background colour down a little bit. I use #EEEEEE for Musica Viva. It's a very light shade of grey that seems to be perceived almost as white but is less tiring to the eye.
Debs wrote:
Funny you should mention orange, that's the color (however very pale ... I call it orange cream) that I am considering on my site.
Have you considered Lightsalmon or Peachpuff? These are light orangy tones dampened by hint of reddish-brown. _________________ Frank Nordberg
http://www.musicaviva.com
The internet center for free sheet music
Funny you should mention orange, that's the color (however very pale ... I call it orange cream) that I am considering on my site. Hopefully your sunglasses will be here by the time I get it online
Sunglasses apart, I was serious about orange and optimism.
Frank wrote:
I was in a fortunate position to do these colour tests. With around 20000 daily page visits I could get statistically relevant numbers pretty fast and since I have a "regular" job I love, risking the ads income from my site for a period didn't bother me that much.
You were doubly fortunate, by the sound of it...
Frank wrote:
I ought to mention that my ad colour tests didn't show any difference between various basic colours. Red, green, blue, grey, yellow all worked just as well. The effect seemed to be solely dependent on visibility
My colour schemes are based more on "traditional wisdom" - a dangerous thing. My sites are many and small, and I wasn't prepared to spend on the PPCs for so many different small tests (that would have taken quite a while to get the traffic levels).
I considered trying some sort of mass test, but concluded that the results would be too vague (too many other variables). From your experience, I think I'm glad I just "guessed", this time!
Thanks for the thread,
Charlie. _________________ "Before I speak, I have something important to say."
- Groucho Marx
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
Your host: Allan Gardyne. Earning a good living from affiliate
programs since 1998.
FAQ
Search
Memberlist
Usergroups
Register
Profile
Private Messages
Log in
)
I looked over your site and will be taking your testing under consideration for my own site.
