3 {{ upvoteCount | shortNum }}
3 {{ upvoteCount | shortNum }}

Very simple Site Split-Testing (AB testing) with Clickbank

by GeorgR.
2 replies
I am testing clicks/sales with a simple site where i promote a clickbank product.

I have a darker themed page, and i made a clone of the page with a brighter theme.

This simple script i found doing some googling:

<?

= 'landingpage1dark.html';
= 'landingpage2bright.html';
= 'landingpage3whatever.html';

//this is the text file, which will be stored in the same directory as this file,
//count.txt needs to be CHMOD to 777, full privlledges, to read and write to it.
= "count.txt";

//open the txt file
= @fopen(, 'r');
= @fread(, 5);
@fclose();

//see which landing page is next in line to be shown.
if ( >= count()) {
= 1;
} else {
= + 1;
}

//write to the txt file.
= fopen(, 'w') or die("can't open file");
= . "n";
fwrite(, );
fclose();

//include the landing page
include_once();

//terminate script
die();

?>
I named it "index.php".

Every time the site is called it cycles through and displays one of the pages in the list.

Now, with clickbank tracking is a little "meh", the only thing in this case i can do is tracking using the TID variable appended to my hoplinks.
(The TID is the clickbank tracking ID which you can see on your CB stats page, it can be alphanumeric, max 8 characters long. Eg. &TID=bright01
or &TID=dark01. You simple append it to your links on the particular site so you know from what page sales are coming from.

Now, each time someone goes to the page it shows either the dark page or the new brighter one...and in a month or so i will have statistics on clickbank which one sells better.

Maybe one of you might find this useful. I know there are more complex tracking systems out there, but i wanted it simple and free.

G.
#clickbank #simple #site #splittesting
  • Profile picture of the author RGallowitz
    I wonder if there would be any significance in difference with the test results?
    I have never really tested background colours (just stayed within the norm). I do however see big chances when testing headlines.
    Signature
    Make INSANE money by promoting PHYSICAL affiliate products.
    The one and only "GALLO Affiliate System" -
    >> Click Here! <<
    {{ DiscussionBoard.errors[1501766].message }}
    • Profile picture of the author whateverpedia
      Originally Posted by RGallowitz View Post

      I wonder if there would be any significance in difference with the test results?
      I have never really tested background colours (just stayed within the norm). I do however see big chances when testing headlines.
      In fact you could use something like this to tweek everything on your sales/landing page down to the the last comma, t and dotted i.

      First tweek to find the best headline, then each papragraph of your sales copy. Colours, images, positioning, you could tweek every pixel on your site for maximum effect.

      It'd take a bit of time and effort, but at the end you should have one HOT sales page.
      Signature
      Why do garden gnomes smell so bad?
      So that blind people can hate them as well.
      {{ DiscussionBoard.errors[1502325].message }}

Trending Topics