MTM tags allow us to collect data in an ‘Event Value’ field and three layers of hierarchy:
Event Category
Event Action
Event Name
In Matomo we can view Event counts at the same three levels. Levels 1 and 2 act as a folder for the deeper layer. At each level, Event Values are shown next to the event count.
As I’m tracking a custom value, I expected that the Event Value field is the best way to store it. However, I find that in the tables under Events the values are always summed for all the event counts. If I’d track the event ‘it took this many seconds to complete the order’ 10 times, I’d get a sum of those 10 durations. Although that makes it easy to calculate the average, it makes it hard to analyze the extremes.
Instead, I could change my tag and fill the Event Name field with the value. That way I can browse through the folders and see a list of values for the Event Action in my tag. That list however, doesn’t give me the option to further analyze the values. It doesn’t provide an easy option to copy or export the names (and counts) as a list either.
How is one supposed to collect numerical values with MTM and Matomo?
How is one supposedo analyze numerical values—is there a way to put them into graphs other than through goal conversions?
Background: Our website contains an iframe. I need to know how long it takes to load. I’ve set up a trigger for a custom event and a tag to do just that. But now I don’t know how to for instance find the 95th percentile and see if long load times are are a common issue or not.
you’re right, the event value field sums up the values, which can be a bit challenging for extreme value analysis. using the event name field is a clever workaround to see a list of values but might bot facilitate further analysis easily
to tackle this, you might consider exploring Matomo’s custom dimensions or variables. they allow you to store and analyze data in a more structured manner, making it easier to work with numerical values.
regarding graphing numerical values, you can create custom graphs based on these custom dimensions or variables. this way, you can visualize load times and easily spot patterns or outliers
included a missunderstood. The Event Tracking function is not 3-dimensional. Its only 2x2-dimensional. Its a little bit tricky to understand it and use it right.
The Category can ignored in the most cases, btw. filled with a placeholder. Action and Name are importand.
With using a iframe it will be comolicated with the transfer of the time from the main document to the iframe document. You can use url parameters for that. Also add the main document start time (Linux Timestam) as a url paramater in the iframe source <iframe src="https://www.example.com?documentstarttime=125874257742" .... And than calculate the time lag and put it in “Event Name”.
Event Category = iframe
Event Action = iframe-timelag
Event Name = value
Thanks for pointing me towards the custom dimensions, Joana. I wasn’t aware of these yet. In case somebody’s interested: described how you can set them up. TBH I found this additional abstraction layer and the additional setup required for the reports a bit too much for this purpose. I’ll just export my values to a spreadsheet every now and then.
Melbao, you’re right—I don’t know what you mean with “The Event Tracking function is not 3-dimensional. Its only 2x2-dimensional.” I couldn’t find documentation on it either. Could you elaborate on that?
Ah okay! I give my tags a category, action and sometimes a name, so in that sense the name is at least in my mind connected to a category, via the action. Doesn’t really matter I guess. The way I see it is that we can categorize actions by feature/topic/whatever and add a name only when we need to store a value related to the action.
