Histogram of Propsum Value "duration"

  • 7 December 2022
  • 3 replies

  • Conversationalist
  • 5 replies

Dear community, I have the following question:

We have created an event property named duration according to this description in order to measure the time that users spend in our app on specific content types.
Now we would like to create a histogram in order to see the distribution of the value of those, but summed up, meaning that I need a PROPSUM distribution chart where I would see how many users spent how much time in our app in a given time period. I found this demo chart that comes close to it, but only shows the distribution of the values at a certain time and not the SUM of those. 

Is there any workaround to get there? I know I could do this with the Session Chart maybe if I set start and end event differently and select length distribution, or maybe the funnel chart with Time to Convert but neither looks like a very clean way to do it. Would be great if anyone could help, many thanks in advance! 



Best answer by Yuanyuan Zhang 12 December 2022, 15:50

View original

3 replies

Userlevel 5
Badge +5

Hi @Asja

I am sharing the email communication here so that the community is aware of the outcome. 

After checking with the team, we concluded that there is no perfect workaround to achieve your use case at the moment. One of my colleagues thought of a workaround: you can sum and group by User ID and then download the breakdown table as a CSV and do the analysis somewhere else like Google Sheets. 



@Yuanyuan Zhang Many thanks for your response and thank you for asking your team about a possible solution. 

I managed to do what you suggested with Excel :) But of course it would be much nicer for us to have that in flexible format in Amplitude so we could build cohorts from our most active user buckets! Maybe this can be considered as a feature request? Many thanks. 



Userlevel 6
Badge +9

Hi @Asja

Glad to hear that you have managed the use case suing Excel. 

I have gone ahead and submitted the feature request for you. Thank you for the valuable feedback!