How should 'view count' of 'point in a video' be counted?

Question

Scenario: There are multiple points in a video that are annotated. User has the list of annotation. User can open any annotation; the annotation will open in video and the view count for that annotation would increment by 1. The view count for annotation is important to showcase popular annotations

Problem: On a video of duration 5 mins. There are 2 annotations A1 and A2 added at time-frame 2 min and 3 min respectively. User opens A1 from the list. View count of A1 increase. User continues watching video and moves pass 3 min annotation A2.

Should the view count for annotation A2 also increment?

• Yes, because user has viewed A2 also
• No, because user's intention was to view A1 and only A1's view count should increment

As an user, What would you expect in this scenario?

Answer 1

It's very difficult to answer this question, because before knowing the how? we need to know the why?. Remember: at any time, at every point, in UX context is everything.

For example: why do you need to track views and or clicks?

This is the paramount question, and knowing the answer to this question will probably give the answer to your specific question, which is just a subset of the main one.

Let's take a look a different scenarios:

• You need to track clicks

In this scenario, you need to track a focal point of the video, and how the user interacts. Then you'll obviously need to track the click, the view is anecdotic since you have no way to know if the user effectively watches the video.

• You need to track views between lengths

In this case the system has to assume that any annotation contained between video_length(0)...video_length(n) has been watched. If you look for more accuracy, you can add a control at the end of the video to know if the user watched the video or if the video was abandoned playing, although this is not very likely

• You need to accurately know what part of the video has been seen

In this case, you can do something like segmenting the video lengths based on annotations. Thus, if user started on A1, when video gets to A2 it should stop and require user interaction. Then, if user chooses to continue, you'll know that s/he chose to watch both segments and intentionality is clear, thus you can count both A1 and A2.

• You only need to track video views, whether partial or complete

In this case (the most common one), then you don't need to worry, simply count how many times the video has been played and that's it

• You need to take track of the point where the user chose to enter the video

Here you will only need to track the click on the annotation

---

These are just a few examples of possible approaches, keep in mind that your case could be within these parameters or be something totally different (for example, for more scientific methods you'll need additional control mechanisms). The point is: as long as you know the why? you'll be closer to the how?

Answer 2

I would count both and also keep an additional parameter capturing whether this view was intentional or forced, alternatively you can capture both and also capture annotation data on click

Answer 3

I think that A1 only should be increased because we know the user intentionally clicked on the A1 annotation on the list, and that we don't know for sure that he has seen the A2 during the video.

So maybe you should keep track of two counters :

• user click view who would keep track of user click on annotations

• video views, which would keep track of the views that the annotation have each time the video goes through it.