Forum Moderators: martinibuster

Message Too Old, No Replies

What's a good test?

         

Fish_Texas

11:16 pm on May 17, 2007 (gmt 0)

10+ Year Member



How long should a test be after changing ad placement, color, size, etc?
How long does it take before Adsense recognizes the change and adjusts the CPC rates up or down?

Feedback from your testing please...
Thanks Fish Texas

Quadrille

10:33 am on May 18, 2007 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Depends on your site.

If you have a million visits a day, then changes should show an effect over about 9.45 hours ;) - but if you have 50 visitors a day, then you'll need to give it a month or more.

nomis5

10:52 am on May 18, 2007 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



At 10,000 a day I seem to see an almost instantaneous effect. I judge the effect by looking at the eCPM for the new or changed adblock. I'm not clear why a million visits per day (as noted in previous post) would take hours.

Quadrille

11:04 am on May 18, 2007 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Note the smiley face and the exact number of hours.

I was merely pointing out that it varies with visitor numbers.

nomis5

11:16 am on May 18, 2007 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



I did see the smiley! But it interested me what you said. Even at 50 visitors a day wouldn't the effect be noticeble within a day? Or is there some way that at 50 a day Google would serve the ads differently to a million a day site.?

Quadrille

11:42 am on May 18, 2007 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



With a small number of visitors a day, you would see a big vaiation in click patters. Some days zero; others quite high.

With a busier site, or a number of days, the total number of visits would tend to even out the 'chance' effects, enabling you to see the 'real' effects of your changes.

Even with a very busy site, a small change in Adsense design might give misleading results if you sampled too small a time frame.

there's also the possibility - with every site that gets repeat visitors - that a change will encourage clicks 'for changes sake', and in the longer term, the previous design may have been a better bet.

So - unless you change frequently and use the chnage to prod your regulars, it may be better to let the new style 'bed down' for a while before making a comparison.

Fish_Texas

11:59 am on May 18, 2007 (gmt 0)

10+ Year Member



Thanks for your feedback, (On this site, visitors are less than 1,000 a day, all SE organic, 20% return visitors...Change made 2 days ago)
I do notice a much higher eCPM and CPC in the morning, but by the end of the day it's back to where I started before the changes.
Perhaps I need to let the dust settle before I draw any conclusions.
Fish Texas