Results 1 to 8 of 8

Thread: YouTube stats of "Audience retention"

  1. #1

    Default YouTube stats of "Audience retention"

    I like the new look of the Channel pages on YouTube and spent a few mins re-looking at the statistics pages.

    AFAIK, anyone can look at the graph of the "Total Views" for any clip. Simply click on the "graph" icon.
    However, if it is *your* clip, a further link appears "View more statistics". The gives an overview page with a link to "Audience retention" and a graph of the percentage of hits against the duration of the clip. In effect, one can see how many of the 'total views' actually saw the entire clip.
    As a consumer, I would prefer to see that graph for anyone else's clips (instead of the rather useless 'Total Views' statistic). But others may prefer the world did not know how most visitors had only watched a few seconds of their masterpieces.

    Of course, one single graph does not tell much of a story. And sadly, without the full web logs, it is impossible to clearly interpret Visitor activity. For example - what was the Retention rate amongst Unique visitors; or the Retention when the Traffic source was a YouTube search.

    Fun to see, but best not taken too seriously. Or is it?

  2. #2

    Default

    When I discovered this stat I thought it would be useful to know at what point of the video most people left so I could see what it was about it and not make the same mistake again. However I have so few views of my videos that YouTube just reports "There were not enough views for the selected date range and content. If possible, please broaden your date range or select different content". So it looks like I need to promote my whole channel for this to be of any use, especially as it's now "monetized". Fat chance of me being able to retire on my YouTube arnings.

  3. #3

    Default

    I believe the stats only exist for clips with X number of views. Not sure what X is. Maybe 200. I recommend trying on one of your higher viewed clips.

  4. #4
    Join Date
    Feb 2006
    Location
    Surrey, UK
    Posts
    10,842
    Blog Entries
    1

    Default

    The relative retention stat has been available for some time, Tim - certainly since longe before thnew layot, but I believe the absolute retention is new.

    I've struggled and failed to work out when the stats kick in. Cetrainly it seems stats only started being gathered at some point between October 2009 and Jan 2010 - anything of mine befiore that comes up with the "not enough views" message.

    I have some more recent with as few as 196 views and stats are available!

    I even have one for which relative retention is available but not absolute retention - which seems totally ilogical given that surey one would need the latter to calculate the former!

    My wierdest disciovery yet is my 4 minute abseil film from last year. According to the stats, absolute retention drops steadily from 100% at 12 secs to 0% at 26 secs where it remains (which might be disappointing, but is feasible), yet the relative retention starts at Low (how can that be at t=0?) moves up to "Below Average" and by 3:22 has crept up to "Above Average". Clearly rubbish as this would mean that at this point other videos must have negative attention.
    Tim

  5. #5

    Default

    Their explanation is here: Audience Retention - YouTube Help and it is worth noting how "fast forward" can affect the stats.
    Is is possible that the clip contained a marvellous section which visitors have been replaying over and over again?

  6. #6
    Join Date
    Feb 2006
    Location
    Surrey, UK
    Posts
    10,842
    Blog Entries
    1

    Default

    Quote Originally Posted by TimAndrews View Post
    Is is possible that the clip contained a marvellous section which visitors have been replaying over and over again?
    LOL (Did I really type that?).
    Thanks for the link, Tim, but it still doen't make sense as repeated playing of a section would also push up the Absolute retention, which remains steadfastly at zero.
    Lies, damn lies and statistics!
    Tim

  7. #7

    Default

    I agree.

    Sadly, we cannot download the raw data. The Relative's sampling rate is based on "videos of a similar length", but they do not define "similar"; or how many other videos were used in the sampling, or how the number of sample periods per minite. I have looked at the graphs for 6 of my clips, but am not wiser. My conclusions were somewhat obvious.

    Kinks and unnatural bends in the graph may occur when clips only have a few hundred hits. As the hits increase, both graphs give smoother curves.

    Hits from the 'Traffic Source' of the youtube search results are the most likely reason for having low retention around the 15 second point. The 'Relative' may only confirm that this is perfectly normal.

    Having lots of text in the YouTube title, tags and description will attract larger number of hits. Lucky, if your clip only lasts 15 seconds.

  8. #8
    Join Date
    Feb 2006
    Location
    Surrey, UK
    Posts
    10,842
    Blog Entries
    1

    Default

    Quote Originally Posted by TimAndrews View Post
    Lucky, if your clip only lasts 15 seconds.
    Hah! Perfect for the Fifteen Second comp that James Webber (Fingercuff) entered "Pants" into. With any sense, the tags will include "underwear" and "nudity". Possibly not what the viewers were expecting.
    Tim

Similar Threads

  1. apparently- "intense" "trippy" "weird"..demo reel
    By thesplashedparietalobe in forum User Videos
    Replies: 13
    Last Post: 05-28-2012, 09:34 AM
  2. "Publish to YouTube" missing in Sony Vegas Pro 10.0?
    By SenTnel in forum Sony Vegas video editing apps
    Replies: 3
    Last Post: 11-17-2010, 08:59 PM
  3. My youtube "rrichiee's REEL time" lifestyle
    By rrichiee in forum YouTube (and others) channel pimpage
    Replies: 3
    Last Post: 02-07-2009, 11:10 PM

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •