Under feed mode, the "Dwell Time" metric is less meaningful, as it may include time when the post was off of the screen. Therefore, I propose that we do one of the following:
1) Introduce a new metric, "Visible Time", that only includes the amount of time where the post was on the screen. This could be calculated as the time where at least 50% of the post is visible, to make it more meaningful.
2) Update the "Dwell Time" metric to only include time where the post was on screen.
3) Make both of these changes, with a subtle difference: The dwell time metric could count the amount of time from when the post was first visible, to when it was last visible. If the user scrolled down so the post was no longer visible, and then back up, this time would be included. Conversely, the "Visible Time" could count only the time where the post was visible.
Under feed mode, the "Dwell Time" metric is less meaningful, as it may include time when the post was off of the screen. Therefore, I propose that we do one of the following:
1) Introduce a new metric, "Visible Time", that only includes the amount of time where the post was on the screen. This could be calculated as the time where at least 50% of the post is visible, to make it more meaningful.
2) Update the "Dwell Time" metric to only include time where the post was on screen.
3) Make both of these changes, with a subtle difference: The dwell time metric could count the amount of time from when the post was first visible, to when it was last visible. If the user scrolled down so the post was no longer visible, and then back up, this time would be included. Conversely, the "Visible Time" could count only the time where the post was visible.