06-02-2004, 10:43 AM
|
|
|
Confirmed User
Join Date: May 2002
Location: European Union
Posts: 3,815
|
Quote:
Originally posted by jayeff
Basically agree with you, but lots of problems. For example, you have Silvercash on your list and they have a ton of sites. I doubt you are sending to all. So your results speak less for Silvercash than for the specific site(s) you are promoting. But we don't know which they are.
The method of promotion is another of a long list of variables. You write your own text links, use stock banners, or...? We have all experienced lousy results with sponsors other people swear by (even allowing for suspected hype and special deals) and vice versa. There is also the reality that when ratios are even as low as 1:100 (never mind the poorer ratios that are common), sales are going to cluster and that makes for unreliable stats on anything but huge numbers.
I know you recognize all this, because I have followed your other threads. So these comments are about the broader issue, not a criticism of what you are doing (except that I would question whether earnings per click is as useful a figure as earnings per impression).
I don't think a meaningful assessment of sponsor values is really possible unless hundreds of webmasters contributed to a single database (even then per site stats, rather than per sponsor, would be more relevant). If someone managed that, the fluctuations due to webmasters' individual methods would start to even out. Otherwise, if there were dozens of lists like yours, seeing who consistently appeared as good or bad could be useful.
|
Just look at my sig  |
|
|