Forum Moderators: martinibuster
Anybody else experience a similar dip?
thx
For several months now, this drop has been reported, mainly because advertisers tend to lower or suspend their CPC ammounts near the end of the month.
Within a few days, you should see a change. If not, look at other reasons for the drop.
Traffic has increased since my rise in the SERPs this week, and that has helped my Sales increase and CTR too.
So i say bring the Christmas shopping on now, NOV & DEC should see more record highs for sales and Adsense publishers.
SN
Some people's absolutely worst day in AdSense history is often another persons best AdSense day in history.
At least, the discrepancy between Sept and Oct's growth has been narrowed down from -30% in the second week of Oct to a more acceptable -15%.
unfortunately, it did not. so we had to bite the bullet and shift to the leaderboard to the homepage. that's when we recovered. but we were holding off using the leaderboard in our homepage for aesthetic and branding purposes.
Friday was pretty bad, though my impressions were down as well. OTOH, the last day of Sept was quite good.
Ditto here. I think a more fair comparison would be to compare the last Friday in October to the last Friday in September rather than the last day in September, which was a Tuesday. In our case, at least, Tuesdays are always significantly better than Fridays.
For fun I do the following {take into consideration I am a half-geek financial guy}:
- I go to the history page and copy the all time history for the account.
- I paste the history in excel.
- I then group the results in groups of seven days. {for example the first day through the seventh day that I ran adsesnse makes group #1}
- I can then graph trends in visits, CTR and earnings over the life of the account. That will let you know trends, not 3 days.
P.S: I also run summary stats, standard deviations and correlation tables {yes I am half nuts}, and of course the highest determinant of earnings is not visits {correlation coeficient of less than .60} it is # of clicks {correlation coeficient of about .82}. To do the later you need some kind of statistics package as doing it in excel is not easy.
That's great for two reasons!
1 - What an interesting way to become aware of AdWords bargains (decrease in AdSense payouts in a topic).
2 - Talk about making lemonade when you life hands you lemons, nice conversion of a bummer into opportunity.
Good job!
thx
Dave.
Just a thought.
(That's pretty much what I expected--the same thing happens every month!)
I wonder if this has to do with post-Halloween effect...
SN
doing it in excel is not easy
davewray exhorted:
Come on everyone, brag about your geekness!
I found the the linear correlation coefficient of:
impressions to earnings: 0.61
CTR to earnings: 0.73
clicks to earnings: 0.85
Which agrees nicely with what you report, loanuniverse.
I graph a seven-day and 30-day moving average on the same axes as my daily. I'm pleased to report the 30-day line is increasing monotonically.
My normalized standard deviation of earnings is 0.46 (approximate translation: about 68% of my days are within 46% of average, or alternatively 32% are more than 46% away from average--I have a quite wide spread in my results). My normalized standard deviation of impressions is 0.26, meaning my earnings vary more widely than my impressions. My normalized standard deviation of clicks is 0.50 (and I'll let you draw your own conclusions from that).
To a least pay some lip service to the topic, the last three days for me were 161%, 113%, and 67% of the seven-day moving average. They were 180%, 136%, and 81% of the 30-day moving average.
Six weeks of data were abused in the making of this post.
