The graph described in this article is displayed in the Views section (tab) of the dashboard.
The Scene views graph shows the drop-off rate between the scenes in a video. The drop-off rate, also referred to as the abandonment rate, represents the percentage of viewers who abandon the video at a certain scene and do not proceed to the next scene. The drop-off rate can be affected by several factors, all of which will be explained in this article.
As you view the graph, note the following important points:
|Like all the dashboard graphs and widgets, the data in the graph is based on the date period that was selected.
|The data for the graph is captured only from video views using the SundaySky player.
|The data in the graph shows the total number of views for each scene.
The bars in the graph represent all the scenes in the last published version of the video (for the date period selected).
The tooltip by Scene views displays the exact date of the video version represented in the graph.
Each bar is composed of two stacked rectangles. The gray rectangle represents the number of potential views while the blue rectangle represents the actual views.
Potential Views answers the question: What is the maximum number of views that this scene can have based on the number of times that the scene can be viewed in the scene line-up?
|If the number of scene views matches the number of potential views, the gray rectangle is not displayed. This is the case when everyone who has the potential to watch a scene actually watches it.
Depending on the viewing activity, the Scene views graph can differ in its visual representation. Following are several examples that you may encounter: Standard Linear Activity, Dip in Potential Views, and Sequential vs Skipping.
The graph above shows a standard drop-off rate as viewers progress naturally through the scenes of the video. The number of potential views decreases with each scene as a result of viewers abandoning after the previous scene.
In the graph above, the number of potential views for the third scene (Device Features) is much lower than that of the second scene. A dramatic dip such as this can be explained by either of the following events:
|The scene is being skipped for some audiences and therefore is not included in the video for all the viewers. Note that although the number of potential viewers is low (36), this scene was watched by almost all of them (35).
|The scene was added after the video was originally published. For example:
1. The date range selected is January 1, 2023 - March 1, 2023.
2. The video was published without the third scene (Device Features) on February 15, 2023.
3. The third scene (Device Features) was added only on February 23, 2023 and the video was then re-published.
4. As the new scene was exposed for a relatively short time during the selected date period, the number of potential views is lower. We can assume that most of the viewers watched the video without the new scene and only a small number of them watched the video after the scene was added.
This graph format is displayed by clicking the Sequential VS Skipping button above the graph. The button is active only if there has been at least one skipping action in the video.
In contrast to the previous two examples where the viewing progression is relatively sequential (linear), this graph also addresses non-sequential (skipped) viewing.
In this graph format, a scene that is re-watched or viewed by skipping to it directly is counted separately.
In the example above, the last scene (Application) was reached both sequentially (blue rectangle) as well as by skipping to it directly (dark pink rectangle). This explains why the last scene has more views than the previous one. In the SundaySky player, a scene can be reached directly by moving the scrubber on the timeline or by using the chapter menu (if enabled).
A scene can also accumulate more views if it is re-watched during the same session.
A bar in this graph format can include a maximum of four stacked rectangles, as illustrated in the legend in the example below: