I'm trying to learn more about how customers search on an ecommerce website. It would be useful to know the distribution of multi-token queries - a simple way to do this would be to count the number of spaces that appear in the term.
For "air fryer" - that's two words separated by a space, so I'd track that as 2 Keywords. "Nintendo Switch Games" would be 3, and so on.
Ideally I'd plot this on a histogram to show the distribution.
Solved! Go to Solution.
Topics help categorize Community content and increase your ability to discover relevant content.
Views
Replies
Total Likes
Try this: use a list prop to track your search term. Reference for list props: https://experienceleague.adobe.com/docs/analytics/implementation/vars/page-vars/prop.html?lang=en#li...
For example, let's say you have configured prop13 as a List prop and track your search terms to it. Assuming you have configured the list delimiter as " " (i.e. space), then tracking the following:
s.prop13 = "Nintendo Switch Games";
should result in getting back the values
"Nintendo"
"Switch"
"Games"
when you run a report on prop13 values.
Take note that with this approach, you're still subject to the 100-byte limit of props. So you can't track "Nintendo Switch Games featuring Mario and Luigi and Bowser who are racing through the streets while throwing mushroom bombs" because that exceeds 100 bytes. If you think you need to track long strings, then you'll have to use a "list" variable https://experienceleague.adobe.com/docs/analytics/implementation/vars/page-vars/list.html?lang=en. Note, though, that each report suite can only have a maximum of 3 "list" variables.
Views
Replies
Total Likes
@sosulli use a custom javascript code in Adobe Launch and to count the number of values that user is typing to search and capture in an eVar and report it out in workspace
Thanks VaniBhemarasetty - is there a way to do this downstream in the report suite? In terms of how our teams and processes are set up, going with custom JS is going to take a long longer to achieve the same results.
Views
Replies
Total Likes
Dear @sosulli ,
If it is possible to create a RegEx Pattern for the said requirement, then you can do it using Classification Rule Builder without the tagging effort.
Thank You, Pratheep Arun Raj B (Arun) | NextRow Digital | Terryn Winter Analytics
However, with this approach, you'll need to know in advance the number of Classified values that you want, which may be a difficult task considering that users are free to enter as many search words as they want.
For example, if you have 2 Classified values only, then "Nintendo Switch Games" would get classified into "Nintendo" and "Switch" but not "Games".
Views
Replies
Total Likes
The solution given is not to classify the values, but to count the spaces to understand the number of words. Yes, the numbers have to be manually entered but it is the better option compared to the implementation.
Views
Replies
Total Likes
Try this: use a list prop to track your search term. Reference for list props: https://experienceleague.adobe.com/docs/analytics/implementation/vars/page-vars/prop.html?lang=en#li...
For example, let's say you have configured prop13 as a List prop and track your search terms to it. Assuming you have configured the list delimiter as " " (i.e. space), then tracking the following:
s.prop13 = "Nintendo Switch Games";
should result in getting back the values
"Nintendo"
"Switch"
"Games"
when you run a report on prop13 values.
Take note that with this approach, you're still subject to the 100-byte limit of props. So you can't track "Nintendo Switch Games featuring Mario and Luigi and Bowser who are racing through the streets while throwing mushroom bombs" because that exceeds 100 bytes. If you think you need to track long strings, then you'll have to use a "list" variable https://experienceleague.adobe.com/docs/analytics/implementation/vars/page-vars/list.html?lang=en. Note, though, that each report suite can only have a maximum of 3 "list" variables.
Views
Replies
Total Likes
Views
Likes
Replies