STA Technical Support FAQ

Answer: The problem you're facing is a cookie problem.

The cookie isn't being read on your thankyou page.

This can have two main causes.

1. The cookie is never set on your landing page.

2. The cookie is set on your landing page, but not read on your thankyou page.

#1 can have many causes but the most common is that your top code snippet is placed incorrectly, and isn't the very first thing on the page.

Another possibility is that the user is not accepting cookies. Though, if you have, say, more than 4 or 5 of these in a row, you can probably rule out the idea that none of these visitors were accepting cookies, since about 95% of visitors accept cookies.

#2 can have a few causes, and the most common are these:

a. The landing page and thankyou page are on different domains. This is not supported with STA.

Answer: You can track how your specific ads perform by using adcodes.

Setting this up properly requires two steps:

• First, you must decide which variable name you want to use in the destination url of the ad, and then modify your destination urls to use this variable.
  
  For instance, you could use the variable name "adcode", and your destination URL for one ad could be something like: http://www.mysite.com/mypage.php?adcode=1, and the url for another ad could be http://www.mysite.com/mypage.php?adcode=2
• Second, once you decide on a variable name, you must tell STA to look for that variable name. You do that on the project creation page for the project you are sending the ad traffic to.
  
  You can specify up to three variables to track at a time. Typically you will need only one, but in some instances it's useful to track more than one parameter at a time (for instance, if you're using SpeedPPC).

• First, from the main support page, there is a filter that will show you your taguchi results based on particular ads.
• Second, the conversion summary report has an "adcode" setting, and this will tell you how each ad is converting.

Answer: Yes.

Just make your landing page a redirect page, and have STA rotate 6 different redirect commands.

Suppose you want to test 6 different offers to the same traffic. Here's some sample code you could use:

<?php $id=2; include('/my_domain/public_html/sta/topper.php');?>
<?php if ($f[1]==1){header("Location: first_URL");}?>
<?php if ($f[1]==2){header("Location: second_URL");}?>
<?php if ($f[1]==3){header("Location: third_URL");}?>
<?php if ($f[1]==4){header("Location: fourth_URL");}?>
<?php if ($f[1]==5){header("Location: fifth_URL");}?>
<?php if ($f[1]==6){header("Location: sixth_URL");}?>

Answer:

• Influence: This number tells you how important this factor is on this page. A higher percentage means that it's an important factor, and might even merit more fine-tuning in a future test.


• Option: This number tells you which option the row pertains to.


• Actions/Impressions:These numbers tell you how many times the option showed up on a page (impression) and how many times a desirable action resulted (action).


• Conversion Rate Range: The conversion rate is just the actions divided by the impressions. But this number might not represent the long range conversion rate. For instance, if you've only had one impression, and one action, your conversion rate is 100%, but you are unlikely to see this percentage over time. The more impressions you have, the closer you should expect to be to the long term conversion rate. Thus as you get more traffic, the range of the predicted conversion rate will get narrower and narrower.
  
  Here's how you read the number. Example: 3.5% +/- 1.5% -- Your predicted conversion rate is 3.5%, and there's a 95% chance that the long term conversion rate will fall between 2.0% and 5.0% (3.5% + 1.5% and 3.5% - 1.5%).


• 95% intervals: This is a graphical representation of the conversion rate range, plus some other information indicated by colors.
  
  The colors indicate how well the option is doing versus the control option (always option 1). Red indicates the odds that the option will fare worse in the long run. Green indicates how confident you should be that the option will fare better in the long run. The more red or green you see, the more confident you should be of the relative outcomes. A lot of grey means that there's not enough information to tell the difference.


• Beat Control?: This number tells you the likelihood, on a percentage basis, that the given option will beat the control in the long run. When the number is highlited yellow, this indicates that there is a greater than 95% chance that the option will beat the control in the long run. This calculation is based on assumptions that may not play out completely well in the real world, but it is the number everyone uses, and gets things pretty close to right.


• Beat Leader?: This number tells you the likelihood, on a percentage basis, that the given option will beat the currently leading option in the long run. When the number is shown in a light grey font, the option is trailing by a significant margin, and has a less than 5% chance of beating the leader in the long run. This calculation is based on the same assumptions as the above calculations, and the same caveats apply.