Results 1 to 8 of 8
  1. #1
    Full Member
    Join Date
    January 18th, 2005
    Posts
    297
    I have been wondering about this for a long time now. Lets say you have a page with multiple links to a merchant (each encoded with the 1x1 tracking pixel of course per guidelines). Now if a user comes to this page and clicks thru to the merchant site, one click is registered and multiple impressions. Isn't this really throwing off stats?

    Let me try an example. Suppose you have 20 links to various pages of a merchant site on one of your pages. So you have 20 links, 20 1x1 tracking images. A user comes and bam, 20 impressions are registered. He clicks a link and one click is registered. So really your success ratio here is 100%. But in your stats your success ratio is 5% (20 impressions, 1 click). Over time, this can really really mess up your stats.

    What do you guys think?

    Hopefully some of the network folks and AMs will pop in here and comment on this. It would really help me out.

    I would think that the only way to get accurate impressions/clicks stats is to have just one tracking image per page. But I am scared to death of modifying any network-supplied code ...

    Would love to get some thoughts about this .......

    Sal @ PHATBargains.com

  2. #2
    Newbie
    Join Date
    January 18th, 2005
    Posts
    1,037
    I would love to hear from more experienced AMs about this, too. Akiva, I'm lookin' at you.

    -patrice

  3. #3
    Retired Member
    Guest
    Yes it does throw off the stats. If I have a page with two banners from the same merchant then I know to divide the number of impressions by two in order to get a real statistic.

    Retired Member

  4. #4
    Full Member
    Join Date
    January 18th, 2005
    Posts
    297
    Yeah what a pain. Unfortunately I have so many pages with differing number of links. I'll need to pull out that scientific calculator and take a course in logistic regression to get some accurate stats

    Sal @ PHATBargains.com

  5. #5
    ABW Ambassador Akiva's Avatar
    Join Date
    January 18th, 2005
    Location
    New Jersey
    Posts
    3,266
    Sal you are indeed correct and that is precisely why I disregard the impression stat. It does give somewhat of an idea of how you're doing but in general, impressions shouldn't be a major influence on your ad placement decisions.

    Akiva Bergstrom
    Affiliate Partner Manager
    <TABLE BORDER=0 cellspacing=2 width=500><TR><TD> <FONT SIZE=1>EssentialApparel.com
    CJ/6%/120 days/Datafeed/Parasite free
    Email: akiva@essentialapparel.com
    Phone: 973.696.6200 ext 751
    > Join now! < </TD><TD><FONT SIZE=1>SportsFanfare.com
    CJ/10%/120 days/Datafeed/Parasite free
    Email: akiva@sportsfanfare.com
    AIM: akivabergs
    > Join now! < </TD></TR></TABLE>

  6. #6
    ABW Ambassador
    Join Date
    January 18th, 2005
    Posts
    4,423
    I never link directly to a merchant (okay, rarely). This allows alot of flexibility, including only showing the trackign bug during the redirect, giving a somewhere skewed stat the otherway, but a little more worthwhile.

    Chet

  7. #7
    Newbie Affiliate Ian's Avatar
    Join Date
    January 18th, 2005
    Posts
    1,540
    Great question! I used to wonder about this also. I am not a stats expert but I do have a decent stats background so let me take a stab at it.

    Yes, the scenarios you showed will definitely skew your CTR. For my own calculations, I look at each set of numbers and decide (very quickly) to inclue them in my overall calculations or not. Some affiliates take out the tracking code and they are easy to filter out. It is a judgement call and jugdement calls are not always statistically correct. But more importantly (at least for our program), I think CTR's are less important that CR. With most CR calculations, you would use "all clicks" and "number of sales". No matter how many impressions were served up, CR only takes into account users who arrive at your site and the number of unique sales. Mind you, this is also not 100% accurate as most programs will show total clicks and not unique visitors. Total clicks does not necessary equal unique users.

    So in summary: CR is a more accurate indication of how well your site is performing and CTR is a nice to know figure but not as important as CR. Like Akiva, I tend to pay less attention to CTR.

    Of course, when you drill down and want to know which actual link your users are clicking through (ie. using Sub-ID tracking), that is another new scenario in which impression might be more important.

    Hope that helps!

    Ian Lee, M.Sc.

    Internet Marketing Strategist / Affiliate Manager
    EYI, Inc.
    ilee_NO_SPAM@eyi.us
    http://www.eyiproducts.com
    604-596-9766

    Personal Sites: ADS-Links.com | HealthCastle.com

  8. #8
    ABW Ambassador
    Join Date
    January 18th, 2005
    Posts
    1,403
    Todd once posted here: "Who really cares for impressions?"
    I personally don't.

    carneol

  9. Newsletter Signup

+ Reply to Thread

Similar Threads

  1. Tracking Code?
    By mailman in forum Midnight Cafe'
    Replies: 2
    Last Post: November 25th, 2002, 05:27 AM
  2. Tracking Code!
    By mailman in forum Commission Junction - CJ
    Replies: 5
    Last Post: September 14th, 2002, 03:08 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •