Msg#: 4553501 posted 4:42 pm on Mar 11, 2013 (gmt 0)
Background: We originally handled our affiliate program in house by using the ?source=affiliateID at the end of each affiliate's URLs. We then moved to a 3rd party program that came to have a lot of issues and they are not only are they expensive but their customer service reps aren't any help at all.
We are wanting to do it in house again but in testing the ?source= doesn't work all the time. What could be the cause of that? Could it be a firewall issue in house? Is there a better way to do it in house?
Msg#: 4553501 posted 5:13 pm on Mar 11, 2013 (gmt 0)
Rather than pass variables on the URL like that, you could use /aff/###/ (or similar) in the URL then use script just like you do with the variable to parse the affiliate's ID. This way the variables on the URL won't be an issue.
Msg#: 4553501 posted 5:47 pm on Mar 11, 2013 (gmt 0)
It could if you don't take steps to exclude the affiliate pages from the indexes. You'd keep the top level urls that point to the products without the /aff/###/ so they could be indexed by the bots. Then any of the affiliate pages could be noindex. Since I'm not really sure how you're setup nor the relationship with your affiliates and how the site is being used, this is all hypothetical.
Msg#: 4553501 posted 6:00 pm on Mar 11, 2013 (gmt 0)
I see and that would definitely work. Could you explain a little bit more about how you could track using the /aff/###/? Currently what we were planning to do (with the ?source=) is to search the source field for the affiliate ID.
Msg#: 4553501 posted 3:41 pm on Mar 12, 2013 (gmt 0)
My thoughts on the issue are that there may be issues with the variables on the URL on the client side - could be browser security setting, AV program, or who knows what (this is a guess). But if the server is handling them behind the scenes then they're out of the picture. Cleaner URL == eliminates the possibility.
@iteri The ### is just a placeholder in the example and would be replaced by an affiliate ID. You could write the URL any way you see fit - skip the /aff/ and just use the /###/ if you want. As long as you can reliably extract the URL into a usable variable when the user hits it.