|Please help me understand these stats.|
| 7:01 am on Jul 5, 2005 (gmt 0)|
I'm having a bit of trouble understanding my stats. Here are some (hypothetical of course) stats for one of my sites:
day 1: 13 clicks, $0.43 ecpm, $1.14 earnings
day 2: 33 clicks, $0.72 ecpm, $0.92 earnings
Until now I was thinking that the higher the ecpm, the better, but am I wrong about this? If someone is paying $0.72 per thousand, then I should make more per click than the $0.43 ones, right? By those stats I would have thought that day 2's earnings would be much higher than day 1's. Can someone please help me understand this?
| 7:08 am on Jul 5, 2005 (gmt 0)|
That eCPM number will be greatly skewed depending on how many impressions you have for each day. The identical EPC (Earnings Per Click) and actual clicks will have vastly different eCPM if there is a great variation in the actual impressions you serve on each day.
A better determination would be comparing your EPC, which was nearly 9 cents for the first day and 3 cents the second, using your hypothetical stats.
| 7:11 am on Jul 5, 2005 (gmt 0)|
eCPS means earnings per thousand page impressions. You want this value to be high.
| 7:20 am on Jul 5, 2005 (gmt 0)|
earnings per click is the way to go.I was tired of understanding the ecpm,all I want to say is sometimes impressions can be adunits and at times pages, depending on wat you have selected to view.
| 7:43 am on Jul 5, 2005 (gmt 0)|
If you have a standard number of ad impressions per page throughout your website(say, all of your pages have a single ad block with 4 ads), then eCPM becomes a relevant and useful metric. In such a case, the relationship of your earnings to the number of page views you get becomes apparent. This would help you determine if a sudden crash in earnings is due to a reduction in EPC, CTR, or page views.
If you have different ad impressions among your various pages, then eCPM becomes almost irrelevant, and all you can depend on for analysis are the CTR and EPC, as the other posters have pointed out.
| 8:16 pm on Jul 6, 2005 (gmt 0)|
Ok, makes sense now. Thanks!