Forum Moderators: martinibuster

Message Too Old, No Replies

the pattern of low traffic - high EPC and high traffic - low EPC

what is the theory behind

         

newbies

7:23 pm on Sep 18, 2004 (gmt 0)

10+ Year Member



I notice a constant pattern from my stats.

During weekends, my clicks are low while CTR is similar to weekdays, I get high EPC.

During weekdays, my clicks are high (3-4 times of that in weekends), I get 20% less EPC than that of weekends.

My site is not related to activities occur during weekends but is work-related. So there is no possibility that weekends have higher payment ads. Has someone noticed the same pattern?

I think this is what "smart pricing" means: G pays you at higher EPC while your clicks are low to make you happy (at least occasionally), and keep most of the share while your clicks are high.

Freedom

7:25 pm on Sep 18, 2004 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



I have just the opposite. My EPC is down 40 percent on the weekends.

Always.

diamondgrl

9:09 pm on Sep 18, 2004 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



I see half the traffic on weekends as I see on weekdays. Yet I see an 8.6% increase in EPC and a 17.4% increase in CTR. No idea why the difference.

Never_again

11:53 pm on Sep 18, 2004 (gmt 0)

10+ Year Member



Sounds to me like you site demographics are different on weekends when compared with weekdays. Weekend users obviously click on higher value ads than do your weekday users.

div01

12:01 am on Sep 19, 2004 (gmt 0)

10+ Year Member



I see the same pattern and have been scratching my head trying to come up with a reason.

dvduval

12:11 am on Sep 19, 2004 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



I have no discernable pattern. I have had some great weekends, and also some great weekdays. I have had high click thru rates and low payout, and low click thru rates and high payout.

trader

1:24 am on Sep 19, 2004 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Yes there are definately patterns to it.

I wrote a detailed post about daily traffic patterns here [webmasterworld.com...]

elguapo

1:35 am on Sep 19, 2004 (gmt 0)

10+ Year Member



Yes there are definately patterns to it.

... except that one site's patterns may be different from another site's patterns. Each site has its own different characteristics that it is hard to make generalizations across the board. The thing I learned most is to agree that there are differences.

If one says there is a pattern, then I believe it. If one says that there is no pattern, I also believe it! Am not too hard to please, I guess ;o)

trader

6:46 am on Sep 19, 2004 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



True, patterns will often vary greatly depending on the category of a web-site and type of visitors.

However, the reason I think my observations here [webmasterworld.com...] are very significant is they are based on a large number of websites and lots of domains online covering a very diverse variety of misc subjects, both personal goods and services, information sites, and also business sites, across a broad spectrum.

You name it and in all likelihood I have a website targeted toward that category (except no adult or gambling). There are probably not many subjects I do not have covered in one way or the other, most likely directly.

According to mathematicians a minimum of 30 samples are needed for good statistical validty. My observations are based on a much higher number than 30.

newbies

8:03 am on Sep 19, 2004 (gmt 0)

10+ Year Member



Sounds to me like you site demographics are different on weekends when compared with weekdays. Weekend users obviously click on higher value ads than do your weekday users.

Visitors to my site are from all over the world but mainly from the US.

I have noticed this pattern for quite a while. check this post of mine
[webmasterworld.com...]

Freedom

8:13 am on Sep 19, 2004 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



I didn't state my experience completely.

I do have a 30-40 percent drop in traffic during the weekend, but my CTR stays the same. It's my EPC that goes down 40 percent.

According to mathematicians a minimum of 30 samples are needed for good statistical validty.

There is also a formula to determine the necessary sample size needed for statistical analysis of a "population." One needs to have an idea of the size of the population first to calculate how big of a sample size is necessary.