How is numMinutes calculated for tagged events?
Hello,
I'm trying to figure out how to analyze time on task. I have a questions focused on the dashboards metrics and on the data pulled from the API.
Pendo Dashboard:
I want to compare the time on task for a flow of pages, but differentiate it based on whether or not a particular feature was clicked in the funnel.
I have 4 pages with multiple features on each, including text entry fields, toggles, buttons, etc.
Let's say I generate a funnel of
page 1 -> page 2 -> feature D on page 2 -> page 3
What is the "time on task" representing in this flow?
For page 1 -> page 2, does this reflect the total time on page 1, including any time spent clicking on various features, before advancing to page 2?
For page 2 -> feature on page 2, what happens if they click any other feature on page 2 before clicking on feature D on page 2? Do they count as "dropped off?"
For feature D on page 2 -> page 3, does the time on task include any additional feature clicks on page 2 before advancing to page 3? Or does a click on any feature count them as "dropped off?"
Pendo API 'numMinutes' Data:
I found this page explaining how minutes are calculated for page view data. Is this still accurate?
Reviewing my data, I have some features where the number of minutes exceeds the number of feature clicks (e.g., 10 minutes, 1 click). Likewise, there are some features where the number of clicks exceeds the number of minutes (2 clicks, 1 minute). How exactly is numMinutes calculated for features?
Finally, are there any known "best practice" solutions to capturing time on task using data pulled from our API? Would we be able to get more accurate time data by pulling a funnel report, rather than calculating it based on numMinutes from the features/pages data? I'd like to do more advanced modeling with our time data in R if possible.
Thank you in advance!
コメント
Hi Jason! That article that you linked is still accurate in terms of how we calculate time-on-page data. For feature clicks, it works in a very similar way. We bucket activity into 60-second chunks of time, then sum up the number of 60-second buckets that activity occurred in. Here is an example:
In this case, they would have 8 events and 3 minutes of activity during the 12-1 pm time period for Button A. That is because they had events (clicks) on the button during 3 different 60-second time periods. I would expect that often you would see where the number of events exceeds the number of minutes - but I would not expect for the number of minutes to exceed the number of events. If you are still seeing that please submit a support ticket for us to take a closer look.
サインインしてコメントを残してください。