| 3:03 pm on May 26, 2008 (gmt 0)|
Cookies would be your best solution. Stash all the keywords in a cookie, with dates. Dump them into a hidden field in the form when they submit it.
| 10:00 pm on May 28, 2008 (gmt 0)|
You may also want to store the referring search engine as well, that way you can track cross-engine traffic.
| 11:20 am on Jun 3, 2008 (gmt 0)|
Thanks, my following question is, does there exist the term "assist keywords" in PPC, and to what extent is it effective tracking them all..
| 1:27 pm on Jun 3, 2008 (gmt 0)|
Can you explain or define "assist keywords" some more please?
| 1:52 pm on Jun 3, 2008 (gmt 0)|
To what I understood, "Assist Keywords" are the ones that do not actually have converting visitors on to the website. However browzer history of converting visitors reflects that these people have earlier visited the website using some other keywords. Now, these other keywords are "Assist Keywords". I hope, I am able to communicate this..
| 2:48 pm on Jun 3, 2008 (gmt 0)|
Okay, that's very clear. I don't have anything to recommend other than capturing all KW with their dates in a cookie, doing a dump at the time of purchase, and using some kind of regression program or neural net analysis to come up with the contribution of each. Or something much simpler and cruder can be used, like weighting based on time lag and # of intermediate KW.
One of the problems with this analysis is going to be comparing the KW patterns for purchases to the KW patterns of those who didn't purchase. If you can't do that, you can't get a handle on causality at all.
I know that big companies do this all the time and most of them use an ordinary stats program like SAS or SPSS. The data collection shouldn't be too hard. Getting an accurate modeling of the data will be.