I am implementing GA tracking for a big social content sharing platform, where users can use #tags for posts, comments etc. and I am looking for a way to correctly track how users use those tags across the website.
Website owner wants to measure:
amount of posts with certain #tags
amount of #tag mentions in comments under posts (which may be related to other #tags)
amount of upvotes/downvotes on posts related to certain #tags
I was planning to use custom dimension "tag" and send its values whenever any tag-related action occurs, i.e. dim. value "bike" on a post where someone used a #bike tag. However there is a problem - most often actions that occur on the website are related to multiple tags simultaneously, i.e. a post with multiple tags, a pageview for one tag and a comment with another tag etc. Unfortunately, it is impossible to send multiple values for the same dimension at once.
Looking for answers, I found this thread, where the best answer recommends to merge multiple values (#tags) for this dimension into one value, where strings (#tags) are separated by whitespaces. This solution allows to drilldown GA report statistics by using REGEX filtering. However the user must specify the #tag that he will use the REGEX filtering for. Therefore this solution lacks a very important feature - it makes it impossible to straightforward sort GA reports by metrics to get to know about trending popular #tags, that the website owners are not aware of (very important for them).
Can I use Data Studio calculated fields to somehow extract those #tags from the merged strings as a "tag" dimension value and calculate them separately? Or maybe there is an another solution? I will appreciate any feedback, thanks!
Jan
Build a db table for tags- this will give each tag an id. (And make it unique)
Then, Whatever data you want to collect, - can be your columns e.g
ID (primary key) | tag | post name | post id | Post category | user id | num upvotes | Timestamp | etc
Then, when running your reports -query the dB however you want to get which ever metric you want.